Search is not available for this dataset
article
stringlengths 4.36k
149k
| summary
stringlengths 32
3.35k
| section_headings
listlengths 1
91
| keywords
listlengths 0
141
| year
stringclasses 13
values | title
stringlengths 20
281
|
|---|---|---|---|---|---|
The regulation of stem cell proliferation in plants is controlled by intercellular signaling pathways driven by the diffusible CLAVATA3 ( CLV3p ) peptide . CLV3p perception is thought to be mediated by an overlapping array of receptors in the stem cell niche including the transmembrane receptor kinase CLV1 , Receptor-Like Protein Kinase 2 ( RPK2 ) , and a dimer of the receptor-like protein CLV2 and the CORYNE ( CRN ) pseudokinase . Mutations in these receptors have qualitatively similar effects on stem cell function but it is unclear if this represents common or divergent signaling outputs . Previous work in heterologous systems has suggested that CLV1 , RPK2 and CLV2/CRN could form higher order complexes but it is also unclear what relevance these putative complexes have to in vivo receptor functions . Here I use the in vivo regulation of a specific transcriptional target of CLV1 signaling in Arabidopsis to demonstrate that , despite the phenotypic similarities between the different receptor mutants , CLV1 controls distinct signaling outputs in living stem cell niches independent of other receptors . This regulation is separable from stem cell proliferation driven by WUSCHEL , a proposed common transcriptional target of CLV3p signaling . In addition , in the absence of CLV1 , CLV1-related receptor kinases are ectopically expressed but also buffer stem cell proliferation through the auto-repression of their own expression . Collectively these data reveal a unique in vivo role for CLV1 separable from other stem cell receptors and provides a framework for dissecting the signaling outputs in stem cell regulation .
Co-operative receptor kinase function is a common feature in both animal and plant signaling systems . Receptor kinase mutants are frequently genetically additive in plants but the molecular mechanisms underlying this effect are often different . For instance , double mutants between the EF-TU RECEPTOR and FLAGELLIN SENSITIVE2 receptor kinases display enhanced susceptibility to bacterial infection above each single mutant [1] , reflecting differences in pathogen derived ligands , followed by quantitative activation of common downstream outputs . On the other hand , additive genetic interactions among mutants in SOMATIC EMBRYOGENESIS RECEPTOR-LIKE KINASE family co-receptor kinases in response to specific ligands reflect quantitative redundancy as co-receptors [2] . Dissecting the molecular basis of redundancy in gene families in plants is often also complicated by unequal contribution from distinct genes and often requires in vivo analysis of signaling outputs or component interactions [3] . Balanced stem cell production in shoot ( SAM ) and floral meristems ( FMs ) is mediated by cell-to-cell signaling pathways initiated by the CLAVATA3 ( CLV3 ) peptide ligand , a founding member of the CLE family of peptides [4] . Mutations in CLV3 lead to excess accumulation of stem cells in both SAM and FMs [5] . CLV3p is thought to be perceived by a series of overlapping receptor kinases which signal to dampen stem cell production . To date , much of the analysis of these receptor-ligand mutants has been at the gross morphological level . Mutations in the different proposed receptors vary considerably in their strength and genetic interactions . It is unclear if the morphological similarities , strength differences , or genetic interactions are due to co-operative cross talk , convergence on a common signaling output , or divergent pathways . CLV3 is secreted from stem cells in the growing tip of the meristem , proteolytically processed and modified to a 13 amino acid diffusible glycopeptide ( CLV3p ) which diffuses broadly throughout the SAM [6–10] . Genetically , CLV3p perception is mediated by the transmembrane receptor kinase CLV1 [11–13] , but also by a heterodimer of the receptor like protein CLV2 and the transmembrane pseudokinase CORYNE ( CRN , [14–17] ) . Additionally , the receptor kinase RECEPTOR-LIKE PROTEIN KINASE 2 ( RPK2 ) may also function in CLV3p perception [18] . All four mutants are resistant to exogenous CLV3p treatment to different degrees suggesting that they act as receptors for CLV3p in vivo . Previous data using overexpression in differentiated tobacco leaf cells has suggested that CLV1 could form higher order complexes with CLV2/CRN leading to the hypothesis that they may signal co-operatively in the SAM [19–22] . However conflicting results have been obtained by different groups and it is not clear if such complexes form in SAM tissues , or when receptors are expressed at endogenous levels in the appropriate cell types . While CLV3p has been reproducibly demonstrated to bind the CLV1 ectodomain [12 , 23] , differing results have been obtained for the ability of CLV2 to bind CLV3p [20 , 23] . In addition , these studies have not tested for potential co-receptor binding of CLV3p . As such it is unclear how , or if , these receptors control stem cell proliferation co-operatively in vivo . Strong alleles of clv1 , such as clv1-4 and clv1-8 , contain missense mutations in the LRR domain of CLV1 [11] . The residues effected in these mutants are highly conserved among related LRR-RKs and structural and biochemical studies with PXY/TDIF receptor ligand pair have shown that these residues direct ligand binding [24–28] . In contrast , clv1 null mutants are significantly weaker [29] . The weak stem cell defects in clv1 null mutant plants can partially be explained by the compensatory up-regulation of CLV1-related receptor kinases BARELY ANY MERISTEM1 ( BAM1 ) , BAM2 and BAM3 [28] . In wild type plants , BAM receptor expression in the center of the SAM , where CLV1 is highly expressed , is undetectable . Consistent with this observation , triple mutants in bam1 bam2 bam3 have no defects in stem cell regulation on their own . In clv1 mutant SAMs BAM receptors are ectopically expressed in the center of the SAM and partially compensate for clv1 . However it is unclear why null clv1 alleles are only weakly compensated by ectopic BAM expression . It is also unclear why strong clv1 alleles are phenotypically more severe . Previous work has suggested that strong clv1 mutant receptors may interfere with CRN/CLV2 signaling [16] . Alternatively , strong clv1 mutant receptors have been suggested to interfere with BAM signaling [30] . It is not clear how this relates to the feedback regulation of BAM expression by CLV1 . CLV1 , and the other putative CLV3p receptors , are proposed to negatively regulate WUSCHEL ( WUS ) expression in the center of the SAM [13 , 31] . WUS is a homeodomain transcription factor and de-repression of WUS in clv3 mutants is thought to drive excess stem cell proliferation [31 , 32] . Despite this , the expression of WUS is robust and co-incident with CLV1 in wild type plants in the center of the SAM and WUS levels do not change dramatically at the cellular levels in response to loss of CLV1 signaling [11 , 28 , 32 , 33] . Unlike WUS , CLV3p-CLV1 signaling fully represses BAM expression in the center of the SAM in wild type plants [28] . Plants expressing up to 300 fold higher levels of CLV3 have a wild type appearance , suggesting that repression of WUS is most effective outside of the physiological range of CLV3p concentration [34] . Interestingly , expression of CLV1 from the WUS promoter is necessary and sufficient to fully complement both clv1 null mutants and clv1 bam1 bam2 bam3 quadruple null mutants back to wild type levels of stem cell regulation [28] . As such , CLV1 operates exclusively in WUS expressing cells of the SAM , despite WUS being a target for transcriptional repression . It is not clear where in the SAM other proposed CLV3p receptors function or if WUS-mediated cell proliferation is linked to BAM transcriptional regulation by CLV1 in vivo . Here I use quantitative genetics , and the highly specific transcriptional repression of BAM3 by CLV1 to demonstrate that CLV1 signals independent of CRN , CLV2 and RPK2 in response to CLV3p in vivo . In clv1 null mutants , ectopic BAM receptors compensate for CLV1 but also act in an additional feedback loop to dampen their own expression in the SAM and buffer stem cell proliferation . Strong alleles of clv1 specifically interfere with this process and have no impact on CLV2/CRN function . Despite their proposed ability to repress WUS , CRN/CLV2 function exclusively in WUS expressing cells of the SAM like CLV1 . Consistent with this , WUS-induced stem cell proliferation is genetically separable from BAM3 regulation by CLV1 . My data demonstrate that despite the qualitative phenotypic similarities , CLV1 signaling outputs diverge from other receptors , and from WUS , and support a model in which CLV1 is functionally independent of other proposed receptors in vivo .
In order to determine the functional relationship between the proposed CLV3p receptors I used previously published null alleles in CLV1 , CLV2 , BAM1 , BAM2 and BAM3 in the Col-0 background [28 , 35] . To date there are no null EMS generated alleles of CRN in any ecotype and no T-DNA insertions in the CRN coding sequence . I therefore used Cas9 to target CRN and create a null mutant . I created a gRNA targeting the signal sequence encoding region of the CRN gene and used the pCUT series of Cas9 vectors to create indels in the CRN gene in the Col-0 ecotype [36] . One of these , hereby referred to as crn-10 , introduced a single thymine base between bases 20 and 21 in the CRN CDS . The mutation introduces a frameshift in the protein after amino acid 6 in the 33 amino acid signal sequence , leading to three in-frame stop codons four amino acids downstream . No other in frame methionine residues are present in , or before , the predicted CRN transmembrane sequence . Thus , the crn-10 allele retains 6 amino acids out of the original 402 in CRN and creates an early stop codon in the CRN signal sequence and deletes all downstream domains of CRN . crn-10 was segregated away from the Cas9 transgene for all subsequent work . crn-10 plants are qualitatively similar to other published crn alleles [16] , with no new phenotypes noted . Wild type FMs give rise to stem cell populations that support stereotypical numbers of floral organs , culminating in the production of two central fused carpels . In clv mutant FMs , the enhanced rate of stem cell production results in increases in floral organ numbers , providing a quantitative measure of stem cell defects [29] . crn-10 plants displayed increased carpel number with a strength comparable to existing clv2 null alleles in Col-0 ( Fig 1A ) . As such crn-10 behaves similarly to recessive non-null EMS alleles of crn such as crn-1 in the La-er ecotype [16] . crn-10 was fully complemented by a CRN-2xmCherry fusion protein expressed from the endogenous CRN promoter ( 44/44 lines , S1A Fig ) as expected [17] . Previous work aimed at testing the hypothesis that CRN encoded a pseudokinase demonstrated that expression of a CRNK146E mutant protein , which further mutates the conserved active site lysine in CRN , fully complements crn-1 when expressed from the native CRN promoter [22] . At the time crn-1 was the only allele available and encodes a protein with a missense mutation in the CRN transmembrane domain ( G70E ) [16] . It is formally possible that if CRN homodimerized , the crn-1 and CRNK146E proteins could cross complement . I therefore transformed crn-10 with a pCRN::CRNK146E-2xmCherry transgene . Again I observed full complementation supporting the previous designation of CRN as a pseudokinase ( 26/26 lines , S1A Fig ) . Recent work has suggested that phosphorylation of CRN at serine 156 is important for function and that S156A substitutions fail to complement crn-1 when expressed from the 35S promoter [37] . In contrast , expression of either a S156A or a S156D CRN variant fully complemented the crn-10 null mutant when expressed from the CRN native promoter ( 26/27 and 24/24 line displaying full complementation for the S156A and S156D CRN variants respectively , S1A Fig ) . The reason for the different complementation results are unknown but could reflect either the use of different crn mutant plants or different transgene promoters in the complementation experiments . I previously demonstrated that CLV1 function in WUS expressing cells is necessary and sufficient for stem cell regulation [28] . Both CLV2 and CRN are expressed broadly in inflorescence tissue ( see S1B Fig , [15 , 16] ) but it is not clear if they function in WUS expressing cells of the SAM like CLV1 . I therefore tested the ability of CLV2-myc and CRN-2xmCherry fusion proteins to complement their respective null mutants when expressed from either their native promoters or the WUS promoter . Like CRN , expression of CLV2-myc from the native CLV2 promoter in clv2 null mutant plants ( rpl10-1 , [35] ) fully complemented stem cell defects in the majority of lines ( Fig 1 ) . Both CRN and CLV2 expression from the WUS promoter also fully complemented their respective null mutant plants in the majority of T1 lines ( Fig 1 ) . Fully complemented lines contained no flowers with more than two carpels , a level of complementation equivalent to complementation of the clv1 bam1 bam2 bam3 quadruple provided by the pWUS::CLV1-2xGFP transgene as previously reported [28] . Collectively these data indicate that like CLV1 , CRN and CLV2 function exclusively in WUS expressing cells of the SAM and FM . My data demonstrate that CRN , CLV2 and CLV1 all function exclusively in WUS-expressing cells in the center of the meristem , however , this observation does not address if they act together at the biochemical level to converge on similar signaling outputs . In wild type plants CLV1 represses the expression of the related BAM receptors in the center of the SAM in response to CLV3p [28] . I therefore asked if CRN and CLV2 participated in CLV1-mediated repression of BAM3 in the SAM center . For simplicity , BAM3 expression was analyzed since BAM1 , BAM2 and BAM3 are all targets of CLV1 in the SAM center , but BAM1 and BAM2 display expression in the SAM epidermis and floral primorida which is clv1-independant [28] . I introgressed the previously characterized pBAM3::Ypet-N7 transgenic line [28] from Col-0 into the null alleles of clv2 and crn , and isolated homozygous transgenic lines in each mutant background . For rpk2 , a CRISPR null ( rpk2-cr ) was generated directly in the homozygous pBAM3::Ypet-N7 wild type transgenic line , and segregated away from the Cas9 transgene for analysis ( see Materials and Methods ) . The pBAM3::Ypet-N7 reporter generates a tandem Ypet fusion protein that is targeted to the nucleus . I then compared the expression of the BAM3 reporter in the L3-L5 cells of the SAM center ( see S2 Fig for image calibration ) . CLV1 is expressed strongly in L3-L5 cells in wild type , clv3 , and clv1-8 meristems and expression in these cells is sufficient to account for all stem cell regulation by CLV1 [25] . Consistent with previous imaging , BAM3 reporter expression was undetectable in the center of the SAM in wild type plants , but was robustly detectable in clv3 mutants and strong alleles of clv1 ( Fig 2A ) . Similar results were found in FMs ( Fig 2A ) , consistent with CLV3p-CLV1 repression of BAM3 expression in both meristems [28] . In contrast , BAM3 was not expressed in the center of SAMs or FMs in either clv2 or crn mutants ( Fig 2A , see S2 Fig for imaging calibration ) . BAM3 is expressed in phloem lineage cells independent of CLV3-CLV1 signaling [28 , 38] . In all plants examined , BAM3 reporter expression was observed in the phloem linage cells of the vasculature outside of the SAM , consistent with previous work [28] , demonstrating that lack of signal in the SAM was not due to reporter silencing in any one line ( for example see S4 Fig ) . Similarly , no ectopic expression of BAM3 was observed in the L3-L5 cells from SAMs or FMs of rpk2 null mutant plants Fig 2A , see Materials and Methods for construction of rpk2 null mutant ) [18] . These data demonstrate that CLV1 signals to repress BAM3 in response to CLV3p independent of CLV2 , CRN , and RPK2 . I sought to test this observation genetically by creating higher order receptor mutants . Owing to their repression by CLV1 in the center of the SAM , BAM receptors do not normally participate in stem cell regulation leading to an invariant two carpels per flower in bam1 bam2 bam3 triple mutants as in wild type Col-0 plants . However , when ectopically expressed in the center of FMs and SAMs in clv1-101 null mutants , BAM receptors partially compensate for the lack of CLV1 and correspondingly clv1-101 bam1 bam2 bam3 mutants greatly enhance carpel numbers of clv1-101 null mutants and display massive SAM over-proliferation during vegetative growth ( Fig 2B and 2C , [28] ) . crn null mutants are phenotypically weaker than clv1-101 null mutants ( Fig 2C ) . However , unlike clv1-101 , crn null mutants were strictly additive with bam1 bam2 bam3 triple mutants in FMs and crn-10 bam1 bam2 bam3 plants were identical to crn alone ( Fig 2C ) . In addition , crn-10 bam1 bam2 bam3 displayed a bam1 bam2 bam3 vegetative SAM phenotype , and lacked the unregulated SAM over-proliferation seen in clv1-101 bam1 bam2 bam3 plants ( Fig 2B ) [39] . This observation is consistent with CRN being dispensable for CLV1 mediated regulation of BAM expression and signaling in vivo . I previously assessed BAM3 reporter expression in strong alleles of clv1 [28] , using the clv1-8 allele in Col-0 which contains a D295N mutation implicated in CLV3p binding [11] . Strong alleles of clv1 are weakly dominant negative and the molecular basis of this remains unclear [29] . Previous work has suggested strong clv1 mutant receptors could be interfering with CRN [16] , or perhaps BAM1 and BAM2 [30] . To test these possibilities I first examined BAM3 expression in the clv1-101 null allele in Col-0 . BAM3 reporter expression was considerably lower in L3-L5 cells of clv1-101 null SAMs compared to both clv3 null and clv1-8 strong alleles ( Fig 3 and Fig 2A , S3 Fig ) . In some clv1-101 null plants , BAM3 expression was nearly undetectable in the SAM center . Expression of BAM3 in FMs was more reliably detected in clv1-101 null plants , but never approached the levels seen in clv3 null and clv1-8 strong alleles ( Fig 3 , Fig 2A ) . Unlike the low levels of BAM3 expression in clv1-101 null plants , BAM3 reporter expression in clv2 , crn or rpk2 was not observed ( Fig 2A , S3 Fig ) . Therefore , while BAM3 expression is de-repressed in the center of the SAM in both null and strong clv1 alleles , the level of de-repression of BAM3 is higher in the strong clv1-8 allele . This result implies that in clv1-101 null mutants , unknown receptor ( s ) signaling is still effective at repressing BAM3 expression . This reduced repression is not due to co-operative receptor function with CRN , as crn clv1-101 double null mutants contained levels of the BAM3 reporter equivalent to the clv1-101 null alone ( Fig 3A ) . Since BAM receptors are ectopically expressed in clv1-101 null SAMs and partially compensate for clv1 they are capable at some level of signaling like CLV1 [28] . I therefore reasoned that ectopic BAM receptors might be dampening their own expression in the center of the SAM in clv1-101 null mutants . To test this I generated bam1 bam2 bam3 and clv1-101 bam1 bam2 bam3 quadruple mutant pBAM3::Ypet-N7 transgenic lines . Consistent with the previous observation that bam1 bam2 bam3 are dispensable for CLV1 function [28] , BAM3 was fully repressed in the center of either SAMs or FMs in bam1 bam2 bam3 mutant plants . In contrast , the weak expression of the BAM3 reporter in the center of SAMs and FMs in clv1-101 null plants was greatly enhanced in clv1-101 bam1 bam2 bam3 quadruple mutants ( Fig 4A ) . This result demonstrates that ectopically expressed BAM receptors in the center of clv1-101 null mutant SAMs signal to dampen their own expression . The level of BAM3 reporter de-repression in clv1-101 bam1 bam2 bam3 plants was comparable to , and occasionally stronger , than in clv1-8 alleles [28] . Previous work has suggested that clv1 missense receptors could interfere with BAM function [30] or perhaps CRN [16] . To test these hypotheses , I generated crn-10 clv1-8 double mutants and clv1-8 bam1 bam2 bam3 mutants . Consistent with the BAM3 imaging , crn-10 clv1-8 mutants were additive with respect to carpel number compared to single mutant plants ( Fig 3B , S5 Fig ) , indicating that clv1-8 receptors do not interfere with CLV2/CRN signaling genetically . If clv1-8 receptors were interfering with the function of unknown receptors , and not BAM receptors , then clv1-8 bam1 bam2 bam3 should be as strong , or stronger , than quadruple null mutants . However , I found that clv1-8 bam1 bam2 bam3 were comparable to clv1-101 bam1 bam2 bam3 quadruple null mutants ( Fig 4B ) . These data support the hypothesis that clv1-8 , and presumably other strong missense clv1 receptors , function exclusively by interfering with the signaling of ectopic BAM receptors in the SAM . This result is consistent with the BAM3 reporter imaging demonstrating that CRN is dispensable for CLV1 signaling , and also implies that like CLV1 , BAM receptors signal in vivo in a CLV2/CRN independent manner . CLV1 signaling strongly represses BAM expression , but has a considerably weaker effect on WUS in physiological ranges of CLV3p [11 , 28 , 33 , 34] . The current model in the field postulates that upregulation of WUS drives the excess cell proliferation in clv class meristems . I sought to test if WUS up-regulation was causally connected to BAM de-repression in the SAM of clv1 mutants . To do this I repeated the experiments in Schoof et al [31] and used the CLV1 promoter to express WUS in the SAM of Col-0 pBAM3::Ypet-N7 plants . In those experiments , ectopic expression of WUS using a two-component inducible system drove SAM enlargement and resulted in flowers with extra carpels , phenocopying clv1 mutants . Of the 36 T1 CLV1p::WUS lines generated in the Col pBAM3::Ypet-N7 background , only 11 plants displayed increases in SAM size and stem fasciation . Despite having enlarged SAMs reminiscent of clv1 SAMs ( Fig 5B ) , no de-repression of BAM3 was observed in the SAM or FMs of any of the enlarged meristem plants examined ( N = 10 , Fig 5A ) . This indicates that WUS-induced over-proliferation of the stem cell niche is genetically separable from BAM3 transcriptional regulation in the CLV1 pathway .
Despite being cloned nearly 20 years ago , we have little understanding about of how CLV1 signals in planta or its relationship to other proposed CLV3p receptors . Here I demonstrate , using BAM3 repression as a readout , that CLV1 signals to control at least some transcriptional outputs independent of CLV2/CRN and RPK2 but fully dependent on CLV3 ( Fig 6 ) . CLV2/CRN have no effect on BAM3 reporter expression by themselves , or in combination with CLV1 , and genetically do not participate in BAM feedback compensation . As such , despite the different receptor mutants having similar qualitative stem cell defects and resistance to CLV3p , CLV1 , RPK2 and CLV2/CRN are functionally separate and converge on distinct signaling outputs in vivo . This result implies that CLV2/CRN are dispensable for CLV3p mediated perception by CLV1 , and that putative CLV2/CRN/CLV1 complexes seen in tobacco overexpression studies are dispensable for CLV1 signaling in vivo . This is consistent with previous data showing that CLV1 traffics from the plasma membrane to the lytic vacuole in response to CLV3p in a CLV2-independent manner [10] , and consistent with additive genetic interactions with clv2 , crn and clv1 [16] . As such , every readout for CLV1 function would suggest that CLV2/CRN are dispensable for CLV3p perception and signaling in vivo . Previously it was suggested that strong clv1 receptors could interfere with CLV2/CRN function in vivo [16] , however I found that there is a significant enhancement of the strong clv1-8 allele in clv1-8 crn-10 double mutants . Genetic analysis and BAM3 repression analysis suggests that the strength of the clv1-8 mutant receptor can be accounted for solely by interfering with ectopic BAM receptors in SAM , supporting previous studies [30] . Despite their independence from CLV1 , clv2 and crn mutants are resistant to ectopic CLE peptide-mediated SAM or RAM termination in several species of plants , and have not been identified in genetic screens for other peptide responses to date . This suggests that there is either a tight relationship between CLV2/CRN and CLV3/CLE ligand function , or that CLV2/CRN impact a developmental process which superficially resembles CLV3p function . Interestingly , while CLV2 and CRN mutants display resistance to CLV3/CLE peptide induced root termination [16 , 40] , they do not display conspicuous root growth or patterning defects in the absence of peptide ligand [16] . I previously demonstrated that mutants that do not function in the CLV3 pathway , but have a higher rates of SAM cell proliferation and an expanded SAM , display resistance to CLE peptide mediated SAM termination [28] . This suggests that CLE peptide resistance per se is insufficient to determine gene function in CLV3p signaling or perception . Based on this , it is formally possible that CLV2/CRN do not function in CLE mediated perception in vivo , as has been suggested by peptide binding assays [23] . The function of CLV2/CRN in CLE mediated signaling , if any , remains enigmatic , but current data demonstrates that CLV1 ligand perception and binding , stability , endomembrane trafficking , and signaling are all independent of CLV2/CRN in planta . clv1 null mutants are phenotypically weak , despite having ectopic BAM expression in the SAM center . However , like CLV1 , BAM receptors also repress BAM expression in the center of the SAM ( Fig 6 ) . As such , clv1 null mutant SAMs contain low levels of BAM expression , potentially explaining why ectopic BAM expression is not sufficient to fully compensate for clv1 . In the organizing center , CLV1 is proposed to also repress WUS expression . Despite this , CLV1 and WUS expression is largely co-incident and expression of either CLV1 or CRN/CLV2 in WUS-expressing cells is sufficient to account for all functions in stem cell regulation . WUS induced SAM over-proliferation can be uncoupled from CLV1-mediated BAM repression . As such , clv1 and clv3 SAMs are functionally different from WUS-induced clv-like SAMs at some level . WUS itself has been proposed to repress CLV1 [33] , but the significance of this is unclear due to the co-incident expression patterns of both genes , the lack of de-repression of BAM3 in enlarged SAMs ectopically expressing WUS , and the fact that uncoupling CLV1 expression from the native CLV1 promoter in SAM cells has no phenotypic consequence other than to complement clv1 null mutants [10 , 28–30 , 39] . In addition , CLV1 mediated repression of WUS is quantitatively different from that for BAM receptors . In the center of wild type SAMs BAM gene expression is nearly undetectable and becomes robustly detectable in clv1 or clv3 mutants . In contrast , WUS is robustly detected in wild type SAMs [28 , 31] . Overexpression of CLV3 represses WUS , however plants expressing up to 300 fold more CLV3 are wild type in appearance [34] . Thus , at endogenous levels CLV3p strongly suppresses BAM expression in a CLV1-dependent manner , but has less of an effect on WUS , which requires considerably higher and potentially non-physiological levels of CLV3p for full repression . Understanding the regulation of BAM expression by CLV1 could lead to new insights into this signaling pathway .
Plant growth and transgenic plant selection was performed as described in [28] . All clv1 , bam1 , bam2 , bam3 and clv2 alleles are in the isogenic Col-0 background and have been characterized previously . Genotyping of plants from crosses were performed using appropriate primers selecting for mutant alleles or T-DNA insertions [28 , 35] . Carpel counts were performed as described with the exception of comparisons using crn-10 or clv2 . In these lines early termination of the SAM and the cessation of flower production was observed after 5 flowers on average as noted before for clv2 mutations in the Col-0 background [35] . Therefore , I counted flowers 6 and beyond for all comparisons using clv2 or crn-10 alleles for all genotypes in those experiments . The transient termination phenotype of crn-10 was not altered by mutations in clv1 , clv3 or in bam1 bam2 bam3 triple null mutants . In bam1 bam2 bam3 clv1 plants stems elongation is highly distorted as is SAM production as described in [28] , making inferences about floral primordia order difficult . As such , all flowers were counted for comparisons between quadruple mutants . Confocal imaging was performed using an inverted Zeiss 710 confocal . Briefly , an inverter adaptor ( LSM Tech , Etters , PA , USA ) was used to allow upright imaging of shoot meristems when attached to the Zeiss 710 . Details on the configuration of the inverter are available on request and will be published elsewhere . Meristem staging , dissection and mounting were performed as described in [28] , with each presented photo being a mean of eight scan cycles . For each experiment a minimum of 6 meristems were imaged and all imaging experiments were repeated with different plant populations two to four times . Imaging settings for the Ypet channel were kept constant across all experiments , but the gain on the red channel for propidium iodide was altered to account for staining differences as necessary . Image settings were calibrated to capture the dynamic linear range in most plants . At these image settings YFP signal in clv1 bam1 bam2 bam3 quadruple mutant SAMs was occasionally saturating in some nuclei ( S2 Fig ) but no signal was detectable in wild type , clv2 , crn or bam1 bam2 bam3 SAM at these settings ( S2 Fig , Figs 2–5 ) . The CRN promoter binary vectors used were described in [17] . For the CLV2 promoter binary , 1250 bp and 588 bp of the 5' and 3' promoter and UTR regions were amplified from Col-0 and fused together using recombinant PCR to introduce a unique BamH1 site and cloned into pBJ shuttle vector . A Gateway cloning cassette was inserted into the BamH1 site and the entire promoter cassette was transferred into the pMOA33 binary vector as a Not1 fragment [41] . Col-0 CLV2 CDS was amplified using primers that allowed cloning into the pENTRD ( Invitrogen ) vector and fused with an in frame MYC epitope to the C-terminus . This was then recombined into either pMOA33 CLV2p or pMOA33 WUSp . The pMOA33 WUSp vector was described as in [28] . For the generation of pMOA33 CLV1p::WUS , a pENTRD WUS CDS clone was recombined into the pMOA33 CLV1p vector used previously [10 , 28] . The pCUT vector system was used to generate the crn-10 allele and rpk2-cr allele used here [36] . Briefly , this vector series co-expresses nuclear targeted Cas9 from the UBQUITIN10 promoter and gRNAs from the U6 promoter . A 20 base pair gRNA target site ( bold ) , including upstream G and downstream PAM ( underlined ) was selected gaagcaaagaagaagaagaaatgg near the initiator methionine codon in the CRN genomic signal sequence encoding region . This target site was used to generate a gRNA that was cloned into pCUT3 as described in [36] . Kanamycin resistant plants were selected in the T1 . In a couple of T1 plants , branches were observed with flowers which all contained elevated levels of carpels relative to wildtype . Seeds were collected specifically from one of these branches and sequencing in the next generation revealed that these branches arose as somatic bi-allelic sectors containing either an A or T insertion ( bold ) at the same location downstream of the initiator ATG ( uppercase ) ( ATGaagcaaagaagaagaagTaaatgg ) leading to equivalent truncated CRN proteins with only 7 amino acids remaining in the signal sequence of CRN ( MKQRRRRKWMstop ) . Plants with the homozygous T insertion event were selected as this provides resistance to Hph1 digestion using the dCAPs primers CRN-10 F gtagaagcagcaatgaagcaaagaagaaggtg and CRN-10 R gttgaagttgtggataagtg [42] , and segregated away from the pCUT3 transgene . Complementation analysis was performed using vectors described in [17] . For rpk2-cr mutant creation , a tandem array of two different U6 promoter gRNA cassettes targeting the RPK2 CDS were gene synthesized by Invitrogen and cloned into pCUT3 as described in [36] . The RPK2 gRNA target sites chosen were aagattactgctcctggtttgg and tcatggctcttaacattagtgg , with the PAM sequence in bold . This vector was transformed directly into the Col-0 pBAM3::Ypet-N7 line . In the T2 generation from a select T1 line , multiple plants displaying an rpk2 phenotype were identified based on male sterility and extra carpels [18 , 43] . Imaging was performed on 10 plants with an rpk2 phenotype . One line lacking the pCUT3 vector was identified by PCR and backcrossed to Col-0 to maintain as a heterozygous owing to the male sterility of rpk2 mutants [43] . This line was then imaged again in the next generation to confirm BAM3 expression patterns . This line , termed rpk2-cr , contains a +A insertion between nucleotides 229 and 230 and an additional +T insertion between nucleotides 279 and 290 in the RPK2 CDS . This results in the production of an RPK2 protein that contains a stop codon directly after amino acid 76 ( serine76 ) downstream of the signal sequence resulting in the deletion of all LRR repeats and downstream transmembrane , juxtamembrane and kinase domains [43] . | The proliferation of plant stem cells in above ground tissues is controlled by a suite of receptors in response to the CLAVATA3 peptide ligand . Receptor signaling in response to CLAVATA3 prevents over-proliferation of stem cells . It is unclear what the functional relationship is between the proposed CLAVATA3 receptors or if they impact common signaling outputs . Here I demonstrate that CLAVATA1 signals independently of the other receptors kinases to control distinct transcriptional outputs independent of stem cell proliferation . Stem cell proliferation is buffered by a two-step mechanism which transcriptionally regulates receptor levels in the stem cell niche . This mechanism helps explain the strict control of stem cell proliferation and could provide new avenues for improving plant growth . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
]
| [
"biotechnology",
"plant",
"anatomy",
"gene",
"regulation",
"plant",
"physiology",
"plant",
"science",
"stem",
"cells",
"genetically",
"modified",
"plants",
"plants",
"flowering",
"plants",
"flower",
"anatomy",
"genetic",
"engineering",
"genetically",
"modified",
"organisms",
"animal",
"cells",
"stem",
"cell",
"niche",
"proteins",
"gene",
"expression",
"biochemistry",
"flowers",
"cell",
"biology",
"post-translational",
"modification",
"carpels",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"meristems",
"plant",
"biotechnology",
"signal",
"peptides",
"organisms"
]
| 2017 | CLAVATA1 controls distinct signaling outputs that buffer shoot stem cell proliferation through a two-step transcriptional compensation loop |
Several approaches exist to ascertain the connectivity of the brain , and these approaches lead to markedly different topologies , often incompatible with each other . Specifically , recent single-cell recording results seem incompatible with current structural connectivity models . We present a novel method that combines anatomical and temporal constraints to generate biologically plausible connectivity patterns of the visual system of the macaque monkey . Our method takes structural connectivity data from the CoCoMac database and recent single-cell recording data as input and employs an optimization technique to arrive at a new connectivity pattern of the visual system that is in agreement with both types of experimental data . The new connectivity pattern yields a revised model that has fewer levels than current models . In addition , it introduces subcortical–cortical connections . We show that these connections are essential for explaining latency data , are consistent with our current knowledge of the structural connectivity of the visual system , and might explain recent functional imaging results in humans . Furthermore we show that the revised model is not underconstrained like previous models and can be extended to include newer data and other kinds of data . We conclude that the revised model of the connectivity of the visual system reflects current knowledge on the structure and function of the visual system and addresses some of the limitations of previous models .
What are the elements of the visual system in the brain and how are they connected ? A large amount of research in vision science has been devoted to these questions . The most apparent way to answer the question of connectivity is to look at the structural connections between functionally defined areas in the visual system , and these have been studied systematically with experimental tracing studies in rats , cats and monkeys . Because of the invasive and time-consuming manner in which these studies have to be done , results are gathered incrementally and scattered across hundreds of separate research publications . If all these separate results are put together , a model of the large-scale structural connectivity of the visual system can be made . For such a model to make sense however , it should be structured according to an organizational principle . Because there are several possible organizational principles , a number of different models can be found in the literature ( see , e . g . , [1]–[4] ) . However , the model published by Felleman and Van Essen in 1991 [3] has since been accepted as a standard model , and is cited in numerous academic books ( e . g . , see [5] ) and articles ( e . g . , see [6] ) . These models can be compared in Figure 1 . However , a number of limitations of all these models have been identified . We will focus on five limitations , namely that most models are ( 1 ) method dependent ( 2 ) indeterminate , ( 3 ) incomplete , ( 4 ) restricted , and ( 5 ) invalid with respect to latency data . We discuss each of these limitations in more detail below . ( 1 ) The structure of the models is dependent on the method used . All three models mentioned earlier are based on studies using different methods . The hierarchical model [3] is based on the pattern of origin and termination of fibers to and from an area , the models by Andersen Asanuma et al . [2] and the model by Distler et al . [4] are based mostly on tracer data and the model by Zeki and Shipp [1] is used to explain segregation and integration of features of the visual image . If we look at Figure 1 we see that all these models differ , yet they all describe the same system . It seems that the structure of these large-scale models is very much dependent on the kind of data used to constrain the model and the phenomena it is designed to explain , whereas these models should agree as they describe the same physical system . ( 2 ) The models are indeterminate . Hilgetag , O'Neill et al . ( 1996 ) developed an algorithm to generate 150 , 000 candidate hierarchies , all of which agree with the anatomical constraints specified by Felleman & Van Essen [3] in their hierarchical model . Every one of these hierarchies violated fewer constraints than the original Felleman and Van Essen hierarchy . In addition , only two of the considered 30 brain areas were assigned to levels consistently across these candidate hierarchies . From these results , Hilgetag et al . concluded that the constraints ( connection types ) used by Felleman & Van Essen reduce the number of candidate hierarchies insufficiently . In other words , the hierarchical model , and probably all models like it are under-constrained . ( 3 ) The models are incomplete . The connectivity data employed by most of the models of the visual system are incomplete because their connection matrix contains many unknown entries [7] . A large number of possible connections have not been investigated . Even with the additional studies done since then , a large percentage remains unknown . ( 4 ) The models are often restricted . The connection matrices of these models contain only cortico-cortical connections within a single hemisphere . Potentially important connections to and from the contralateral hemisphere and to and from subcortical structures are excluded from the connection matrix [7] . Some of the models in Figure 1 do not even include the LGN . ( 5 ) The models are invalid with respect to latency data . A final limitation identified by Mountcastle [7] concerns the validity of the models in the light of latency data . According to Mountcastle , single unit recordings show that areas in different levels of the hierarchy of these models are nevertheless simultaneously activated by visual stimuli . Clearly , such simultaneous activation seems to disagree with the associated areas being at different levels in the hierarchy . These five limitations question the validity of large-scale models of the visual system . Now , more than 20 years after the inception of the first large-scale models , novel insights and experimental results may give rise to a reconsideration of these models . In this paper , we present a method that integrates the anatomical constraints with functional ( temporal ) constraints extracted from more recent experimental results . Our method leads to the identification of a new structure that meets all the constraints imposed by the integrated data . Before presenting our approach , we expand on the fifth limitation mentioned above , viz . the validity of current large-scale models of the visual system in the light of single unit latency recording studies [8] , [9] . In our method we will use the data from these studies as temporal constraints for our model . First-spike time coding might be the only available neural mechanism to bring about rapid behavioral responses , as the alternative neural code , rate coding , is too slow for such responses ( e . g . , [10] , [11] ) . It has been shown that first spikes can be selective for orientation , faces and optical flow [12] . Given these considerations , an analysis of the spike arrival times at various cortical areas may reveal part of the underlying functional architecture . Schmolesky and colleagues [8] performed a study in which they measured the onset latencies of single-cell responses at several cortical areas in individual anesthetized macaque-monkey brains , evoked by flashing visual stimuli . In Figure 2 , we replotted the results obtained by Schmolesky et al . These data reveal a number of inconsistencies with the large-scale anatomical models . For instance we would expect that first spikes arrive later in areas at higher levels in a hierarchy than those in lower levels , but when comparing the latencies in Figure 2 with the models in Figure 1 it turns out that this is often not the case . For example , we can see that area FEF , which is in level 7 in our schematized Figure 1 ( it is even at level 9 of the original publication [3] ) has latencies comparable to those in V3 ( level 3/4 ) , MT ( level 3–5 ) and MST ( level 3–7 ) . The latencies of FEF even overlap considerably with those of V1 ( level 2 ) . Schmolesky et al . also noted this and they concluded “our data simply indicate that the … anatomical hierarchies fail to account for the initial flow of signals in the visual system and therefore may not accurately represent the ‘functional’ hierarchy of the visual system” ( p . 3277 ) . It seems that large-scale anatomical models of the visual system do not agree with the timing data of Schmolesky et al . [8] . Several others have also noted similar inconsistencies . In a review of mammalian spike timing data , Bair [13] observed that neurons assigned to different hierarchical levels are often activated at the same time or in the wrong order with respect to their presumed hierarchical relation . In another review of the available latency data , Nowak and Bullier [14] state that “latencies to visual stimulation in monkey are not ordered as expected from [anatomical hierarchies]” ( p . 229 ) and they cite several studies in support of this claim . The conduction velocity in the efferent fibers also plays a role in latencies , but this alone cannot explain the discrepancies between timing data and structural data fully and determining or approximating this velocity is not possible with our current knowledge . Taken together , it seems that the response latencies in the visual system are incompatible with large-scale models of the known structural connectivity of the visual system . We aim at resolving this inconsistency by integrating structural connectivity data with single unit recording data . Finding networks with a connectivity pattern that fits both kinds of data might not only eliminate the observed inconsistencies , but also suggest a different model better constrained by multiple sources of data . If the model is more constrained , we might learn more about possible connections that have not been investigated yet , or have been excluded from the model . If the generated networks differ substantially ( in both fit and architecture ) from current models we cannot only conclude that the hierarchical model does not explain the timing data optimally , but we can also point to possible reasons . We want to do this by adopting an extendible method ( it should be possible to add both new data and other kinds of data ) that can use both structural and functional data as “converging evidence” to model the structure of the visual system . The structural data will be extracted from a database , CoCoMac , which combines several hundred studies about the anatomical connectivity of the adult macaque brain . The functional data will be taken from a single representative study and a review article of the available latency data . To find network topologies that fit both kinds of data an optimization method called simulated annealing will be used . A schematic overview of the method can be seen in Figure 3 . In short , with our new method we want to arrive at a revised , more constrained model of the visual system , integrating both structural and functional data , with the additional goal of mitigating the problems of exclusion and incompleteness .
In the first dataset we included the areas used in the study by Schmolesky , Wang et al . [8] . This dataset contains 8 areas: LGN , V1 , V2 , V3 , V4 , MT , MST , and FEF . The LGN is the node of the system that receives the first input , the seed node . Note that because we are only studying bottom-up processing , any subcortical area can play exactly the same role as the LGN in the dynamics of the system . Therefore we will call this node SCA ( Sub Cortical Area ) . As we will see later , subcortical areas such as the superior colliculus and pulvinar are also likely candidates for this role . The connectivity data were extracted from the CoCoMac data base [15] in the form of an 8 by 8 connectivity matrix . The resulting average connection matrix is shown in Figure 4 . New connection entries ( as compared to the original connection matrix ) have been marked with an asterisk . New connections from the SCA have been added to almost every area in the network , except to V4 . ( The connection to V1 was already present . ) In Figure 5 the measures of fit fanat and flat of our resulting network are compared to those of the most influential large-scale structural model , hierarchical model [3] . The connectivity and timing costs of the hierarchical model are compared to the costs of the new resulting connectivity matrix . Our network clearly fits the data better . The structural fit ( meaning the fit of the network with the known structural data ) is maximal ( fanat = 1 ) for the solution . This means that our network does not have any connections that are incompatible with the known tracer data . There is an increase in the temporal fit with respect to the hierarchical model , implying that our network fits the available timing data better than the hierarchical model . Simply put: the newly generated network explains the first spike timing data better without violating any constraint from the known tracer data . Figure 6 depicts our solution superimposed on an actual macaque brain . The newly introduced connections are colored . It is clear that a subcortical route has been added to accommodate the temporal constraints . As we shall argue in the discussion this finding confirms Mountcastle's ( 1998 ) suspicion that subcortical routes play a significant role in cortical connectivity and agrees with recent connectivity studies . In our results with dataset 1 we have shown that the method works for a small dataset from one parcellation scheme . With dataset 2 we want to use our method on a larger dataset , with more areas from different parcellation schemes ( using ORT , [16] ) , hoping to increase the validity of our claims . Also , a larger dataset might increase the constraints on the model . The first spike time data chosen for this purpose were those of Lamme & Roelfsema [12] . After applying the inclusion criteria , we used the data for the following 27 areas: SCA , V1 , 7ip , V3 , Ipa , Pga , TE2 , TE3 , 5 , 7a , FEF , FST , MST , MT , SEF , SMA , V2 , V4 , MI , TS , TAa , TE1 , TEa , Tem , and TPO . The resulting connection matrix can be seen in Figure 7 . New cell entries that differ from the original connectivity matrix have been marked . The connections added to the matrix as a result of including the constraints of the timing data are again mainly connections from the Sub-Cortical Areas ( SCA ) . New cell values also show that some connections have to be absent in order to account for the timing data . In Figure 8 the fits of the most prominent large-scale model , the hierarchical model [3] , and our network are compared . Again , the connectivity and timing costs of the hierarchical model are compared to the costs of the new resulting connectivity matrix . The anatomical fit is at its maximum value of 1 , meaning that the resulting network is in complete concordance with the known anatomical data . Our network is clearly an improvement over the hierarchical model regarding the fit with the latency data ( an increase from 0 . 67 to 0 . 93 ) . These results indicate that adding sub-cortical routes to the network yields a substantial increase of the timing fit . These results establish that the inclusion of the subcortical routes is also essential for explaining latencies in large visual networks . As mentioned in the introduction , one of the shortcomings of the large-scale structural connectivity models is their indeterminate organization , most notably the indeterminate organization of the hierarchical model [17] . Our method integrates two types of constraints , thereby reducing the “indeterminateness” . To confirm the reduction of indeterminateness , 100 solutions were analyzed to determine the levels at which individual areas were placed . Subsequently , we determinen the average level ( and variance ) of each area . Figure 9 shows the results of this analysis . In the left-most part of Figure 9 we can see that when using our method with anatomical constraints only , we get roughly the same indeterminate organization as observed for the hierarchical model , even though we used a different cost function than was used in the original publication . Areas can be at multiple levels without violating the anatomical constraints as Hilgetag , O'Neill et al . [17] have pointed out . The right part of Figure 9 shows the results for the combined temporal and anatomical constraints . The number of candidate hierarchies is reduced considerably . More areas are confined to a single level of the hierarchy , and the remaining areas have a smaller number of possible levels to which they can be assigned . Overall the determinateness ( or constrained-ness ) is greatly improved due to the inclusion of timing data .
One of the strong points of the method used is that it is completely data driven . Large-scale network architectures can be generated with one main assumption only: the time it takes for a signal to pass from one area through the afferent pathway to another area can be treated as a unitary whole . We do not need to rely on any assumptions about the structure or the hierarchy of the areas in the network . One of the possible pitfalls of neural networks models is that the complexity of the model does not add to the predictive power of the model . Sometimes models require setting many parameter-values of which the validity is hard to ascertain . In our method there are only four annealing parameters that need to be set to perform an analysis , the effect of these parameters are well-known [18] and the results do not depend critically on the specific parameter settings , as we can see in Figure 10 . See Methods , Optimization , for more details . Our method combined two types of constraints derived from connectivity and timing data , which is more than the single type of constraint used in earlier work [1] , [3] , [19] . One might argue that more than two types of constraints may or need to be included to generate even more plausible models of cortical connectivity . We mention three additional types of constraints: ( 1 ) the conduction velocity , ( 2 ) the inter-area distance and , ( 3 ) variance in spike time latencies . Although the inclusion of additional types of constraints is rather straightforward in our method , we also motivate why we did not include them in our method . Do our results really solve the inconsistencies between the latency data and the earlier anatomical models ? Clearly , if a model agrees with latency data , all the areas assigned to higher levels of the hierarchy should have longer latencies then those assigned to the lower levels . In Figure 11 the timing data is plotted as a function of the level an area is in . This is done for the hierarchical model [3] ( Figure 11A and 11C ) and for our network ( Figure 11B and 11D ) . In Figure 11A and 11C the sequence of the first spikes does not follow the levels , once again illustrating the shortcomings of all earlier large-scale models . For instance , in Figure 11A , FEF and V4 pop-out as areas with “wrong” timing for their position in the hierarchy . Figure 11B and 11D show that the networks generated by our method do not suffer from this shortcoming; each area is assigned to its appropriate temporal level . As is evident from Figure 11B and 11D , our solution has three levels only . We have argued that our results solve the inconsistencies between the latency data and earlier anatomical models . An earlier study has attempted to show that these inconsistencies do not exist [20] . Petroni , Panzeri et al . simulated the cortical network of the areas examined by Schmolesky et al . [8] , using the connectivity matrix of the hierarchical model [3] . However , Petroni , Panzeri et al . [20] used something they call the hodology [29] , i . e . , the shortest possible route to connect two areas , to determine the level at which an area is placed , instead of the levels of the earlier models [1]–[4] . Their simulated latencies resemble the real latencies . They conclude that hodology correlates better with latency than the hierarchical organization . However , our work shows that even the hodological hierarchy can not explain the latencies fully . In the results of Petroni , Panzeri et al . [20] it is already evident that the timing in areas FEF and V4 , prominent dorsal and ventral areas , do not fit the model . In their own words: “two areas where we found some disagreement between simulated and experimental latency were FEF … and V4” . Furthermore , as we have argued earlier , even a hierarchy based on hodology cannot explain the extremely short latencies or :reversals” in the latencies . Our results show that the hodological hierarchy cannot explain the overall latencies optimally . In Figure 11 we have shown that even when hodology is considered ( the topology we use is always based on the shortest route ) the hierarchical model remains incompatible with the latencies . We shall argue below that our explanation provides a better alternative , because it addresses some of the limitations of previous models and might explain some functional characteristics of the human visual system . For both our datasets , the most prominent and remarkable result is the large increase in goodness of fit with the latency resulting from the introduction of direct connections from the subcortical areas ( SCA ) to the cortical areas , bypassing V1 . Such direct connections are typical long-distance projections . It has been suggested that the brain shows strict optimal component placement and therefore only has short projections between adjacent brain areas . However Kaiser and Hilgetag [28] noted that long-range projections also exist and that they have an essential role to play , e . g . in the minimization of processing steps . Our findings demonstrate this principle perfectly . Additional analysis reveals that discarding all subcortical areas from the dataset , the best fit is to connect all extrastriate areas to V1 , with the exception of V4 . Although the resulting connectivity pattern is still superior in terms of fitness to the hierarchical model ( i . e . , fanat = 0 . 947 , flat = 0 . 871 ) the fits are still smaller than those obtained in our main solution . Apparently subcortical pathways explain the data better then cortico-cortical “shortcuts” . The importance of the subcortical route has been recognized before in various publications . Lamme and Roelfsema [12] noted that a reason for the lack of correspondence between the hierarchical organization and the response latencies might be that subcortical structures like the LGN , the superior colliculus ( SC ) , and the pulvinar ( PUL ) , also project to various extrastriate areas . Even Petroni et al . [20] , who claim there is no large discrepancy between latencies and the hierarchical model , propose that the extremely fast response in FEF might be caused by a subcortical connection through the superior colliculus . It is known that not only the LGN , but also the PUL and the SC have retinal inputs [30]–[32] , making direct connections from the cortex to these areas true shortcuts from the retina . With the additional help of the CoCoMac database [15] and Objective Relational Translation ( ORT ) analysis [16] , a literature search was done to assess the connections from the subcortical areas suggested above , to the cortical visual areas named in our results . Except for the obvious connections to V1 , the LGN is also connected to V2 [33] . The pulvinar is a structure that is densely connected to cortical areas , i . e . , it is connected to V2 [34] , V3 [35] , V4 [36] , MST [37] FEF [38] , and MT [39] . Interestingly , MT not only resembles an early visual area because of this subcortical connection but it is also an early visual area in the way that it matures [6] . Although earlier studies repeatedly suggested the SC might be connected directly to the visual cortex [20] , [40] , little structural evidence for the existence for such connections was found . Relatively few studies have addressed the presence of these connections . A connection from the SC to V1 exists [41] , but connections to the following visual areas were examined but not found: V2 [42] , V3 [43] , PO [34] , IT [44] . A possibility is that the SC is involved indirectly by feeding input to the pulvinar [45] or that it is connected via the mediodorsal thalamus [46] . Summarizing , the subcortical routes suggested by our results are in line with the above findings as all “new” subcortical connections ( connections from SCA in the figures ) found in the results from dataset 1 have been confirmed by anatomical research . These routes run through the lateral geniculate nucleus and the pulvinar ( but not through the superior colliculus ) . The fact that the predictions made by our method have been verified validates our method and lends credibility to our results . Subcortical pathways appear to play a larger role in the propagation of the visual signal then was assumed before . More specifically , the connections from subcortical areas to extrastriate areas have been left out of most models of the visual system . Therefore these connections need to be considered in any future description of the structure of the visual system . Because of all the reasons described above we propose a new , revised organization of areas that is based on both timing and structural data ( see Figure 12 ) . It is important to note that the organization proposed here is based on two kinds of data only , and as such only reflects the combination of first spikes and structural connectivity . Therefore , it can not claim to explain other aspects of hierarchy like the increasing complexity of visual responses or receptive field size . In how far the organization we propose reflects the “general organizational principle” ( if such a thing exists at all ) might be dependent on the importance of the processing of first spikes compared to all processing in the cortex . It might well be that the inclusion of other kinds of data ( e . g . , receptive field size ) will reorder the organization . Another possibility is that in order to explain different kinds of data , different organizations are necessary , and no single organization exists that can explain all aspects of the visual system . Our study is entirely based on connectivity and timing data from the macaque visual system . What does our work tell us about the structure of the human visual system ? Although the homologies between monkey and human visual cortex remain uncertain for some areas , one of the main reasons for studying the monkey visual cortex are the clear similarities with the human visual cortex [47] . All the areas from dataset 1 have a more or less clear homology in the human brain [48] , [49] enabling at least some generalization from our results to the human visual system . Our results , when generalized to the human brain , might explain some recent findings in humans . Goebel et al . [50] looked at the functioning of the dorsal and the ventral stream in two blindsight patients with long-standing post-geniculate lesions ( FS and GY ) . These patients show close to normal brain activity in hMT+ and V4 although a large part of V1 has been destroyed . Similar results were found in a patient with hemianopia in the entire right visual field , who could still report movement and color change in his blind hemifield ( Riddich syndrome ) . fMRI activity was reported in V4/V8 and V5 in the lesioned hemisphere and MEG recording showed it preceded V2/V3 activity [51] . A functional connectivity study showed that there was a flow of information from V5 to V4 and V2 [52] . How can areas higher “upstream” in the visual system be activated normally when almost all of their input from lower levels has been cut ? Subcortical pathways like the ones suggested by our simulations , originating in LGN and the pulvinar might play an important role in explaining the residual functioning of the brain in blindsight and Riddich syndrome . To further understand the structure of the visual system , future work should include attempts to complete the subcortical pathways in the connectivity matrices . This goal might be reached by adding existing tracer studies to the CoCoMac connectivity database [15] or by doing new tracer studies into subcortical pathways . It would also be very helpful for our understanding of the large-scale structure of the visual system to see more latency studies like the one done by Schmolesky et al . [8] with more areas . This would allow us to use the current method without having to rely on the “averaging” methods currently used in dataset 2 [from 12] . Our current approach could also be aided by more research into conduction velocities in the cerebral cortex , the exact role of conduction velocities in explaining latency data remains an open question . A modeling study similar to this one using conduction velocity as an ( extra ) constraint might help to resolve this question . By combining data from both structural connectivity and spike timing experiments using a data driven method with few assumptions and parameters , topologies that fit both kinds of data have been found . The results show the necessity of subcortical routes to explain spike-timing data . Review of the literature demonstrates that most of the connections predicted by our method appear to exist . Furthermore we show that we are able to further constrain our model , in effect reducing the problem of indeterminacy associated with previous models [17] . We conclude that our method successfully incorporates structural and functional data to arrive at a new large-scale model of the visual system that underscores the importance of subcortical routes .
Anatomical connectivity data was obtained from the CoCoMac ( “Collation of Connectivity data on the Macaque brain” ) database [53] . At the moment of writing it contains the details of more then 400 studies about the anatomical connectivity of the adult macaque brain using tracer studies . CoCoMac represents all of this data in an objective , coordinate-free , parcellation-based fashion and enables the user to integrate contradictory findings in the literature , depending on the choice of several parameters . The advantage of using CoCoMac over data from individual studies is that it combines hundreds of tracer studies into a single connectivity matrix . With a mathematical method called ORT ( Objective Relational Transformation ) , it is possible to combine and transform brain mapping data from any parcellation scheme to a coordinate-independent freely chosen parcellation scheme [16] . This allows us to combine connectivity data on areas from several parcellation schemes in one of our datasets ( dataset 2 ) . The database is publicly available and can be queried through the online interface CoCoMac-Online at http://www . cocomac . org/[15] . We define two datasets for the timing data . The first dataset is from a single study mentioned in the introduction [8] . The data was collected from four monkeys , over a relatively large number of recording units ( 558 ) in nine areas of the brain , measured repeatedly and with a broad range of visual stimuli designed to elicit a response from the entire visual system . We obtained part of the original data from the authors and therefore we were also able to determine the variance in the data . The second dataset is from a review which collected data from multiple studies ( [12] , box 2 , p 573 ) . It therefore includes data on more areas then the first dataset . It suits our purposes very well , as it not only includes several studies but it also weighs them as to reflect the reliability of each experimental finding . Comparisons of timing data in the literature often suffer from differences in experimental and analytic methodologies between studies . Although the first dataset is the smaller of the two , it does not suffer from these “incompatibility effects” , because all measurements were made within individual monkeys using common stimulus presentation and analysis techniques . This is especially important as we are interested in differences between latencies across the visual system . The second dataset is larger than the first , allowing us to determine if the results can be generalized to larger systems . However , the data in the second dataset are probably more prone to “incompatibility effects” because they come from a more varied set of experiments and experimental conditions . For each of the two datasets , two conditions needed to be satisfied before the data belonging to a particular brain area could be included: ( 1 ) Both first spike data and connectivity data are available . For instance , the connectivity data for Ts , Ts1 , Ts2 , and Ts3 were available whereas the functional data were only available for Ts . Therefore only Ts was included in the analysis . ( 2 ) Areas included in the selection should be considered part of the visual system and should not be too large to be useful in the analysis . For instance , area PreFr ( prefrontal ) from Lamme and Roelfsema [12] was excluded because it contains a very large number of other areas and because it is arguably not part of the visual system . Because the exact mapping relation between areas FEF and 8a is controversial ( [3] , e . g . , compare [54] , [55] we decided to exclude 8a to resolve any uncertainties . In order to be able to optimize the connectivity matrix , we need a numerical measure of the goodness of fit of timing and connectivity data for any given connectivity pattern . This allows us to search for the best fitting network and is essential for the used optimization technique employed . Below , we define measures for the anatomical fit and temporal fit and combine them into a single overall fitness measure . The definition of the anatomical fit is relatively straightforward . We define the anatomical fit fanat on the unit interval as the proportion of corresponding connections between a candidate connection matrix ( CMA ) and the anatomical connection matrix retrieved from CoCoMac ( CMB ) , i . e . : When fanat is 0 there are no corresponding connections between A and B ( worst fit ) , and when fanat is 1 both are in complete agreement ( best fit ) . Note that the fit will decrease when a connection that is established absent in CMB is present in CMA , but the fit will be unaffected when a connection that is empty CMB is present in CMA . This implies that violating an established absent connection constraint is treated identical to violating an established existing connection . The temporal ( latency ) fit , flat , is defined as a linearly transformed Pearson product-moment correlation coefficient ( PMCC ) . It expresses the similarity between ( the order and magnitude of ) the timing data and our simulated timing data ( see below ) . The function flat should provide an indication of the degree to which the orders of activation of nodes in two networks agree . Although rank correlation seems more appropriate for the simulated timing data because they contain ordinals , the use of rank correlation would mean losing all sensitivity to distances between latencies and therefore PMCC is used because it accommodates the continuous values of the real timing data , which is measured at the ratio level . The minimum value of the PMCC is −1 ( worst fit ) and the maximum is 1 ( best fit ) . Because fanat has a range of 0 to 1 and we want flat to have the same range we apply a linear transformation so that: In order to simulate the timing data for our candidate connection patterns , we employ a version of the “shortest path” algorithm [56] . We start by determining the node that receives the initial input ( the seed node ) . This node is defined as being active at the first time step ( t = 1 ) . We then use the following function to propagate activity through the network for t>1:where ai ( t ) represents the activity of node i at time t and wij the element of the connection matrix , i . e . , the connection from node i to node j representing the presence ( wij = 1 ) or absence ( wij = 0 ) of a connection . Nodes connected to an active node will be active the next time-step and will remain active afterwards ( wii = 1 , for all i ) . If all nodes of the network are active , each node is assigned a level equal to the number of connection steps from the seed node . In all simulations , the number of time steps is sufficiently large to allow activations to propagate through the entire network . The overall fit function F is defined as the weighted sum of both fitness measures , i . e . : The two measures of fit are weighted by factor α . The optimization algorithm used in our method is simulated annealing . Simulated annealing is a stochastic combinatorial optimization algorithm belonging to the class of methods known as gradient descent algorithms . What makes it especially suited for our needs is that it is capable of finding solutions that obey multiple types of constraints . In addition it is generic in that it does not require an explicit knowledge description of the problem at hand [57] . The simulated annealing algorithm was first described by Kirkpatrick , Gelatt et al . [18] and has been applied before for optimizing features of cortical networks [28] . In our method , the algorithm makes changes to individual elements of the connection matrix by applying the rule that each change should increase the fitness . The algorithm varies the strictness with which the rule is applied . Initially , the degree of randomness ( or temperature T ) is high , meaning that changes that decrease the fitness are also allowed , and the values of individual elements are updated at random . As time progresses , the degree of randomness is lowered towards full determinism ( i . e . , zero temperature ) . At this stage , connections are updated by applying the rule strictly . The quality of the final solution ( i . e . , the agreement with the anatomical and temporal constraints ) depends on the rate of change from randomness towards determinism . The annealing schedule defines the rate in terms of the time-dependent temperature T ( n ) :where n is the iteration step of the algorithm and τ is the annealing factor . By allowing changes that reduce fit during early iterations of the algorithm while later making these changes very improbable , the use of temperature allows the algorithm to avoid local minima , i . e . , solutions that fulfill a subset of the constraints but are not the optimal solution . Other gradient descent algorithms could have gotten “stuck” in these local minima because multiple connections need to be flipped to result in an increase of the fitness . The employed method makes it possible to generate networks that fit both timing and connectivity data without any need for assumptions about hierarchy or connectivity . Figure 3 presents a schematic overview of our method . The latency data ( shown on the left ) and the anatomical data ( shown on the right ) are transformed into constraints . Starting from a random connectivity pattern , the simulated annealing algorithm ( shown in the middle ) uses both types of constraints to generate candidate connectivity patterns that obey the constraints . Our method can be used with any selection of brain areas and can easily be extended to include newer data as it becomes available . Other kinds of data can also be added with relative ease as long as a fitness function can be devised for it . We used the following parameters for the optimization algorithm in all our simulations . The initial temperature was T ( 0 ) = 4 and the annealing factor was set to τ = 0 . 99 , the maximum number of annealing iterations was set to 1500 , and the parameter α was set to 0 . 5 so the contribution of timing ( flat ) and connectivity ( fanat ) fit to the total fitness was balanced . It should be noted that the two measures of fit , fanat and flat , are not necessarily equally sensitive to changing the state of one connection . For every solution we ensured that when the algorithm terminates , at least hundred iterations of the simulated annealing algorithm had not lead to new solutions . We repeated all our simulations to show that the results are stable over a wide range of values for the above parameters . We varied the values of α , T ( 0 ) , τ . We also varied the amount of connections in the randomly generated network that serves as the starting point for the annealing algorithm ( initial edge density ) . For α and initial edge density the range could be varied over the entire possible range . The results of these tests can be seen in Figure 10 . Except for the drop-offs at extreme values of α and a few small peaks , the results are essentially stable over the entire range of the parameters . The drop-offs are a result of the fact that at extreme values of α , the networks are fitted to one fit function only ( either flat or fanat ) , instead of to two . The small peaks are sub-optimal solutions that disappear after the averaging of results described in the next section . Because the optimization method described above is stochastic , some form of averaging over multiple solutions is needed . In our method , multiple solutions are multiple connection matrices . All our results are based on 1000 computed optimal connection matrices , each using different initial connection matrices . All the matrices can be summarized into one matrix by defining each element as a probability of the presence of a connection . When a number of solutions are generated by our method , two types of values will be found in the cells . The first type does not change when increasing the number of solutions , and these values are always 100 or 0 percent . The values of 100 and 0 percent represent connections that were present or did not exist , respectively , in all the simulations . The second type has a value that asymptotically moves closer to an intermediate percentage . For instance , the value of 50 percent means that this particular connection is present in half of the solutions and , presumably , does not matter for the fit of this network . These connections are connections that were not constrained by our data; they have not been researched according to CoCoMac [15] and do not matter for the first spike timing in the network ( e . g . , they could represent a feedback connection instead of a feedforward connection ) . When we excluded all the solutions that are not optimal ( as expressed by the amount of total fit with the data ) all elements have values of 0 , 50 , or 100 percent . The resulting connection matrices therefore only contain the values 0 ( 0% ) 1 ( 100% ) or they will be empty ( 50% ) . The computational tools needed to import the data from the CoCoMac database ( Cocomac Import/Export Tool ) and the tools to perform the optimization and analyze the data ( BrainAnnealer ) are custom programs written in Object Pascal with the Borland Delphi compiler . Tools , source and documentation together with the data used can be downloaded at http://www . capalbo . nl . | Visual perception is very important to us , something we can easily come to realize if we imagine ourselves blind . The visual system consists of numerous interconnected brain areas . If we are to understand the functioning of the visual system , then we will need to understand the connectivity between these areas . Current models of the visual system have a number of limitations . One of these is that the time it takes for the neural signal to reach a certain area often seems inconsistent with the place of that area in the overall structure of the system; e . g . , the signal might arrive relatively quickly at an area generally located “higher” in the visual system and slowly at an area located in the “lower” part . We combine data about the known connectivity in the monkey brain with timing data to find a network structure that is consistent with both kinds of data . The results show that the timing data can be explained when the network contains direct routes from subcortical areas to “higher” cortical areas . We show that our model has fewer limitations than previous models and might explain unresolved issues in the study of connectivity in the human brain . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
]
| [
"neuroscience/theoretical",
"neuroscience",
"neuroscience/sensory",
"systems",
"computational",
"biology/computational",
"neuroscience"
]
| 2008 | Combining Structural Connectivity and Response Latencies to Model the Structure of the Visual System |
Cellular morphology and associated morphodynamics are widely used for qualitative and quantitative assessments of cell state . Here we implement a framework to profile cellular morphodynamics based on an adaptive decomposition of local cell boundary motion into instantaneous frequency spectra defined by the Hilbert-Huang transform ( HHT ) . Our approach revealed that spontaneously migrating cells with approximately homogeneous molecular makeup show remarkably consistent instantaneous frequency distributions , though they have markedly heterogeneous mobility . Distinctions in cell edge motion between these cells are captured predominantly by differences in the magnitude of the frequencies . We found that acute photo-inhibition of Vav2 guanine exchange factor , an activator of the Rho family of signaling proteins coordinating cell motility , produces significant shifts in the frequency distribution , but does not affect frequency magnitude . We therefore concluded that the frequency spectrum encodes the wiring of the molecular circuitry that regulates cell boundary movements , whereas the magnitude captures the activation level of the circuitry . We also used HHT spectra as multi-scale spatiotemporal features in statistical region merging to identify subcellular regions of distinct motion behavior . In line with our conclusion that different HHT spectra relate to different signaling regimes , we found that subcellular regions with different morphodynamics indeed exhibit distinct Rac1 activities . This algorithm thus can serve as an accurate and sensitive classifier of cellular morphodynamics to pinpoint spatial and temporal boundaries between signaling regimes .
Cell morphology and morphodynamics are used to phenotype the state of a cell throughout various processes , including differentiation , proliferation , migration and apoptosis[1–5] . Moreover , numerous signaling pathways converge onto cytoskeleton architecture that determines morphological variation among cells . Therefore , parameters of cell morphology and morphodynamics can also serve as indicators of signaling states[6 , 7] . Indeed , analysis of cellular morphology and morphodynamics has been applied , for example , in cancer cell screens[8] , drug development[9–11] , cell transformation characterization[12] and cell cycle analysis[13 , 14] . A number of strategies have been developed to elucidate the physical cause and signaling regulation of cell morphology . Quantification of cell edge movements using physical and mathematical models revealed different modes of motility associated with actin-based spreading[6 , 15–25] , myosin-related contraction[26] and transverse wave propagation[22 , 27–31] . Moreover , shape descriptors have been used for statistical classification of cell morphological patterns[32–38] . However , these studies generally applied a global parameterization of cell morphological changes , such as degree of polarization , cell area change and migration rate , and did not consider the local and dynamic behavior of the cell edge . This has been in part due to the significant complexities in robustly tracking cell edge motion at the subcellular scale . We[39] and others[40] have developed the necessary image analysis framework to track complex cell boundary movements in time-lapse cell image sequences . Densely sampled protrusion and retraction velocities were compiled in space-time maps that offer an opportunity to identify distinct cell morphodynamic states as well as to unveil putative functional links to underlying cytoskeleton dynamics[41–44] and signaling activities[45–47] . Nonetheless , a systematic classification of the spatiotemporal patterns captured by these maps has yet to be performed . Here , we implement a framework based on the Hilbert-Huang Transform ( HHT ) to decompose the spatiotemporal signal into instantaneous frequencies and amplitudes . Applied to a population of spontaneously migrating fibroblast-like Cos7 cells , we show that the frequencies encode information on the wiring topology of pathways involved in the regulation of morphodynamics , whereas the amplitudes reflect pathway activation levels . We then validate these results by acute manipulation of the wiring topology of a pathway using optogenetics[38] . We also show that the decomposition into temporally and spatially localized frequency spectra offers an opportunity to identify time windows and cell edge sectors with distinct morphodynamic signatures . This permits determination with subcellular resolution of switches between morphodynamic states that are associated with particular signaling motifs .
We hypothesized that subcellular morphodynamic profiling would be highly informative regarding the states of signaling pathways that regulate cytoskeleton and adhesion dynamics at the cell periphery . To test this , we first imaged unstimulated Cos7 monkey kidney fibroblast-like cells . These cells often exhibit a robust spontaneous migration , and because of their tight adhesion to the substrate , are ideal for high-resolution live cell imaging . We tracked the motion of virtual fiduciaries on the cell boundary by identifying the outline of the cell edge in each frame of a time-lapse sequence , and mapping the outlines of consecutive frames subject to minimizing the overall displacement and strain that are associated with the deforming cell shape[39 , 48] ( Fig 1A , see Methods and S1 Fig for details on the mapping strategy ) . We subsequently sampled time series of local protrusion ( positive velocities ) and retraction ( negative velocities ) by averaging the motion within edge sectors of ~10 pixel ( i . e . ~3 μm ) width each . This spacing in sampling corresponds to the half-width-at-half-maximum ( HWHM ) of the spatial autocorrelation of the edge motion[49] . Velocity time series along the cell boundary were then compiled sector-by-sector into the rows of a matrix referred to as a protrusion activity map[39] ( Fig 1B ) . Accordingly , a matrix column represents the velocity variation over all edge sectors in a particular time point . For the particular cell displayed in Fig 1A , the boundary region encompassing sectors 12–38 prominently protrudes for the first 15 min of the movie , interspersed with short periods of retraction . After 15 min the region splits into two protrusive subregions . The boundary region encompassing sectors 40–54 retracts for the first 10 min before converting into a relatively quiescent zone ( see also Video 1 ) . These examples show that the velocity time series is nonstationary . Accordingly , edge motion analysis must be temporally localized . To analyze edge velocity time series , we adopted the Hilbert–Huang transform ( HHT ) [50–55] . The HHT relies on an empirical mode decomposition ( EMD ) , which divides the submitted time series into a finite and generally small number of component signals , referred to as intrinsic mode functions ( IMFs ) . The set of IMFs forms a complete and nearly orthogonal basis for the original signal satisfying the following two conditions: i ) The number of local extrema and the number of zero-crossings either is equal to each other or at most differs by one . ii ) The mean value of the upper envelope defined by the local maxima and the lower envelope defined by the local minima is equal to zero . Under these conditions , the Hilbert Transform is guaranteed to converge to an unbiased estimate of the instantaneous frequency spectrum of the IMF[50 , 54] . The EMD procedure involves iterative application of the following steps: i ) Identifying all local extrema in the original target time series X ( t ) . ii ) Connecting all local maxima by a cubic spline to generate the upper envelope; similarly , connecting all local minima by a cubic spline to generate the lower envelope . iii ) Computing the mean m1 ( t ) of upper and lower envelopes and subtracting it from the target time series to generate a reduced series h1 ( t ) . If h1 ( t ) satisfies the aforementioned conditions of an IMF , it is the first IMF component c1 ( t ) . Usually this is not the case . Instead h1 ( 1 ) ( t ) = h1 ( t ) is considered the new target time series , and the above procedure is repeated k-1 times , h1 ( 1 ) ( t ) −m1 ( 2 ) ( t ) =h1 ( 2 ) ( t ) ⁞h1 ( k−1 ) ( t ) −m1 ( k ) ( t ) =h1 ( k ) ( t ) until h1 ( k ) ( t ) satisfies the conditions of an IMF . This is the first IMF component c1 ( t ) . The residual signal r1 ( t ) is then defined as r1 ( t ) =X ( t ) −c1 ( t ) and used in a next iteration as the initial target time series . Usually , the decomposition is terminated after n iterations , subject to the condition that the residual signal is either a constant , or a monotonic function , or a function with only one maximum and one minimum , from which no more IMF can be generated . However , in our application IMF sets were compared between experiments . Therefore , it was necessary to fix the number of iterations such that the majority of decomposed data fulfilled the above defined termination criterion . Irrespective of the termination rule , the EMD generates n IMF components c1 ( t ) , … , cn ( t ) and a residual signal rn ( t ) that satisfy X ( t ) =∑i=1nci ( t ) +rn ( t ) where rn ( t ) either fulfills the above termination criterion or its variance is less than , e . g . , 5% of that of the original target time series X ( t ) . Application of the Hilbert Transform to a particular IMF produces an instantaneous frequency spectrum at each time point t . H[ci ( t ) ]=1π∫−∞∞ci ( τ ) t−τdτ ( 1 ) F ( t ) =12π⋅ddt ( arctan ( H[ci ( t ) ]ci ( t ) ) ) ( 2 ) A ( t ) =ci2 ( t ) +H2[ci ( t ) ] ( 3 ) where i = 1 , … , n . The instantaneous frequency spectrum is the temporal derivative of the phase change in the IMF signal ci ( t ) , which is defined by the inverse tangent function of the quotient between the Hilbert Transform of the original signal ci ( t ) ( see Eq ( 1 ) ) and the original signal ci ( t ) ( see Eq ( 2 ) ) . The corresponding instantaneous amplitude spectrum is the root of the square sum of the original signal ci ( t ) and its Hilbert Transform ( see Eq ( 3 ) ) . We show an example of the decomposition of the velocity time series at a specific sector ( Fig 1C ) in Fig 1D . By definition of the EMD procedure , higher order IMFs ( Fig 1D ) tend to contain lower frequencies and lower amplitudes . However , in our data the instantaneous frequency and amplitude spectra at a specific sector overlapped between IMFs ( Fig 1E and 1F ) . We computed the frequency and amplitude spectra for all IMFs in all edge sectors , which generated at each time point for each cell boundary sector six temporal frequency and amplitude values . Moreover , we repeated the HHT computation for all columns of the protrusion activity map to capture the instantaneous spatial frequency and amplitude spectra . As with the time domain , we restricted the EMD to six spatial IMFs , which generated at each cell boundary sector for each time point another six spatial frequency and amplitude values . We chose the number ( six ) of IMFs empirically and found it works well to capture the variation of complex cellular morphodynamics . To illustrate the meaning of the EMD and to better interpret the related spectral decomposition outcomes , we reconstructed six movies and associated activity maps that visualize the cell edge motion captured by the six IMFs ( Fig 1D , S2 Fig and and Video 2 ) . For a particular IMF at a particular cell edge sector we extracted time point by time point the velocity magnitude and integrated the values into a displacement time series ( Fig 1C ) . After computing the displacements for all sectors in one time point we plotted the virtual cell edge and repeated the procedure for all time points to generate a movie associated with the IMF . Each of the six movies starts with the true cell edge image at the first time point . Video 2 clearly indicates the distinct levels of motion persistence and magnitude captured by the six IMF signals . For example , IMF1 captured the protrusion signal with highest frequency and greatest magnitude , which yields rapid and jerky changes in cell shape . In contrast , IMF6 captured only subtle long-range position changes of the cell edge with almost no shape change associated . Hence , the instantaneous frequencies extracted from these different IMF orders represent , on average , different length scales and ranges of persistence in the protrusion-retraction cycles of a cell . We first applied the spectral decomposition to the edge movements of spontaneously protruding Cos7 fibroblasts . These cells exhibited a wide range of cell shapes and morphodynamics at a basal level of activity . For example , some cells showed persistent polarity and protruded/retracted over large parts of their peripheries ( top panel in Fig 2A and Video1 ) . Other cells showed an unpolarized morphology with only small oscillatory edge movements along the entire periphery ( lower panel in Fig 2A and Video 3 ) . For the two cells illustrated in Fig 2A , we extracted histograms of instantaneous frequencies from each of the six IMFs ( Fig 2B ) . Despite the vast differences in cell shape and motion , the two sets of histograms appeared strikingly similar . For both active and quiescent cell , the central frequencies of IMFs decreased exponentially ( Fig 2C ) . Comparison of cumulative distribution functions ( CDFs ) using Kolmogorov–Smirnov ( K-S ) test statistics confirmed that the frequency spectra of the two cells were statistically indistinguishable ( Fig 2D and S3 Fig ) . In contrast , the K-S test statistics of the instantaneous amplitudes were different ( Fig 2E and S3 Fig ) . This observation also held for 48 spontaneously protruding Cos7 cells ( Fig 2F and 2G ) . The instantaneous frequency distributions of cells with comparable molecular makeup and similar levels of stimulation were conserved regardless of morphological and morphodynamic differences . In contrast , morphological and morphodynamic differences manifested themselves in significant variations of the amplitude spectra . The more different the velocities of two cells were , the larger the difference between their instantaneous amplitude spectra ( Fig 2G ) ) . Of note , the small differences between instantaneous frequency spectra were independent of the cell order ( Fig 2F ) . Those analyses indicate the orthogonality between instantaneous amplitude and frequency spectra in capturing cell morphodynamic behaviors . The conservation of instantaneous frequency distribution in molecularly similar , spontaneously migrating Cos7 cells led us to ask whether induced shifts in morphogenetic signaling greater than the basal level of variation in a control cell population would systematically alter the frequency components . To address this , we employed a recently introduced optogenetic construct that allows acute and reversible inhibition of the guanosine exchange factor ( GEF ) Vav2[38] . We have previously shown that Vav2 acts as a core element of a signaling resonator that controls the oscillatory protrusion and retraction of cells[56] . To capture the morphodynamic response to Vav2 inhibition , we filmed cells for 6 minutes without light-activation of the inhibitor construct , followed by 12 min of pulsed blue-light inhibition , and then another 12 min in the dark to examine the recovery of Vav2 activation levels ( Fig 3A ) . Activation pulses of three different lengths were examined: 1000 msec , 100 msec , and 1 msec ( Fig 3A ) . Based on the comparison of instantaneous frequency distributions , photo-inhibition with pulse lengths of 1000 msec and 100 msec changed the spectra . We did not observe any evident change with a pulse length of 1 msec ( see Fig 3B ) . Importantly , the shifts were limited to the first three IMFs , which covered frequencies in a range 0 . 006–0 . 035 Hz ( S4 Fig ) . Frequencies below this band were unaffected . Overall , inhibition of Vav2 signaling yielded lower frequencies , suggesting that this signal is implicated in pathways that promote fast exploratory protrusion and retraction cycles . Strikingly , scatter plots of frequency versus amplitude indicate that Vav2 inhibition has no effect on the amplitude ( Fig 3C ) , i . e . the speed of the protrusion-retraction cycles . Those experiments suggest a separation of pathways that set the pace of the protrusion machinery from pathways that define the power of this same machinery . In the experiments described thus far , we used instantaneous frequency and amplitude as morphodynamic signatures reflecting the state of an entire cell with sufficient sensitivity . We then asked if those signatures could also be applied to distinguish the potentially transient signaling states of subcellular regions . We first evaluated the spectral signatures of a migrating Cos7 cell with obvious polarity ( Fig 4A and 4B , upper panels ) . The subcellular region indicated by the solid green box represents the actively protruding cell front , whereas the region indicated by dashed green box represents the retracting/quiescent cell rear . We separately applied the HHT to the time series encompassed by these two regions , extracted the instantaneous frequency distributions and conducted the K-S test to obtain K-S statistics of all six IMFs . For comparison , we repeated this analysis on two randomly selected subcellular regions of a quiescent Cos7 cell ( Fig 4A and 4B , lower panels ) . For all IMFs , the K-S statistics comparing the front to back dynamics in a polarized cell was greater than the K-S statistics comparing two randomly selected regions of a quiescent cell ( Fig 4C ) . The former K-S statistics also systematically exceeded the average K-S statistics quantifying cell-to-cell variability in the population of control cells ( Fig 4C , black dash line ) analyzed in Fig 2 . However , they did not exceed the level of K-S statistics that were related to the morphodynamic shifts induced by acute Vav2 inhibition ( Fig 4C , red dash line ) . This suggests that the signaling changes we experimentally introduced were stronger than the differences in signaling programs between front and back of a polarized cell . To further test the postulation that spectral signatures of cell edge motion could distinguish the signaling states of subcellular regions , we compiled the instantaneous temporal and spatial frequency and amplitude spectra in a feature vector at each time point and each location and performed statistical region merging ( SRM ) [57] to identify regions of the cell edge with distinct motion regimens . Specifically , we formulated two 12-dimensional vectors in each sector s at each time point t ( Eq ( 4 ) ) . The vector contains components 1 to 6 of the instantaneous temporal frequencies at time point t computed from the sector’s six IMFs along the time axis . Components 7 to 12 contain the instantaneous spatial frequencies in sector s computed from the six IMFs capturing the cell edge undulations at time point t along the space axis . The vector represented in Eq ( 5 ) captures the instantaneous amplitudes in the same fashion . The feature vector ϕ ( s , t ) in each sector s at each time point t is then composed of amplitude-weighted instantaneous temporal and spatial frequencies ( Eq ( 6 ) ) . The amplitude weights are normalized by Amax , t ( s ) , which denotes the maximum amplitude for a specific sector along the time axis and by Amax , s ( t ) , denoting the maximum amplitude at a specific time point along the spatial axis . We chose quadratic amplitudes because they reflect the instantaneous relative energy consumed by a particular IMF in the temporal and spatial domain . In summary , the feature vector captures the instantaneous spectral properties that characterize the local morphodynamic activity of a particular sector at a particular time point . We exploited the feature vector to identify in the protrusion activity map regions of homogeneous morphodynamics , i . e . regions of the cell edge that move over a specific time period under the same regimen . To define such regions we applied the SRM algorithm[57] . For a multi-dimensional feature vector , this algorithm merges two regions R1 and R2 if the difference in every feature component between the two regions is less than a threshold ( Eq ( 7 ) ) . The threshold penalizes regions of very large area and includes a user-controlled merging delicacy parameter Q ( Eq ( 8 ) ) . |Rj| denotes the size of a region , and |Rj|max is an estimate for the largest region clustered in the map . Throughout this work , we set the value of |Rj|max to 256 . Nt is the number of time frames in the cell imaging , and Ns is the number of sectors or windows along the cell periphery . The merging started with the feature vectors in individual edge sectors and time points and iteratively grew regions with sufficient similarity in morphodynamics until none of two regions in the protrusion activity map fulfilled the merging criteria . We showed the response of SRM to different levels of merging delicacy in Fig 5A . At Q = 0 , only two protrusion regimens were differentiated , while at Q = 8 the activity map was decomposed into a high number of regimens that spanned very few sectors and lasted for only a few time points . To determine an optimal value for Q , we computed the ratio of intra- vs inter-region variance as a function of Q ( Fig 5B ) . Beyond Q = 3 the fraction of explained variance increased only marginally , indicating that this level of granularity captures the spectrum of relevant morphodynamic regimens . To demonstrate how critical the combination of instantaneous frequency and amplitude is for the formulation of a distinguishing feature vector , we compared the SRM results of the full feature vector using the combined instantaneous frequencies and amplitudes ( Fig 5A ) versus the results from using the instantaneous amplitudes only ( Fig 5C ) . We also computed SRM results using the weighted instantaneous frequency of IMF1 only ( Fig 5D ) . It is evident that the combined frequency and amplitude features accounting for all IMFs captures much finer spatiotemporal patterns . Thus , this feature vector is effective and suitable for SRM clustering . We applied SRM to two cells with distinct initial morphodynamics ( Fig 6A , and Videos 6–7 ) . Both cells were perturbed for 12 min by photoinhibition of Vav2 activity and then released for another 12 min . The first cell displayed a clear polarity with a morphodynamically active front between sector 20 and 50 and a more quiescent back . The difference in this activity is easily perceived in the protrusion activity map ( Fig 6B , top ) and , as with the cell presented in Fig 4 , described by clearly separated motion regimens , where the active front breaks into two regimens ( red and orange ) with slightly different morphodynamic feature values ( Fig 6B , bottom ) . The remainder of the cell edge was described by a single regimen with significantly lower feature values , reflecting the relative quiescence of this cell region . During Vav2 photoinhibition the active front was abrogated and largely merged with the more quiescent regimen . Interestingly , after release from the inhibition the higher activity regimens were restored , yet around the entire cell perimeter . Hence , while the cell regained full morphodynamic activity , it lost polarity . The second cell was less active overall and showed weaker polarity . The effects of Vav2 photo-inhibition were much harder to perceive in the protrusion activity map ( Fig 6C , top ) , yet the region merging unveiled a clear demarcation of motion regimens before and during photo-inhibition ( Fig 6C , bottom ) . After release from inhibition , the cell restored for short time intervals and along the entire perimeter the regimens of the more active zone before inhibition . Together , these experiments highlight the sensitivity of the instantaneous spectral decomposition to outline the spatial and temporal boundaries of distinct morphodynamic activity patterns . Based on our finding that acute switches in Vav2 activity cause acute shifts in the instantaneous frequency spectra of cell edge motion ( Fig 3 ) , we hypothesized that the different motion regimens identified by SRM analysis could be associated with peripheral cell areas of distinct signaling activity . To test this hypothesis , we employed a Förster resonance energy transfer ( FRET ) biosensor probing the activity of the GTPase Rac1 in Cos7 cells , which is one of the targets of Vav2 ( Fig 7A , and Video 8 ) and a key regulator of cytoskeleton processes implicated in cell protrusion activity . Like the construction of the protrusion activity map , we sampled the Rac1 activity locally in probing windows . Each window corresponded one-to-one with a 3 μm–wide sector for protrusion measurements and had a window depth of 3 μm . The average activity values per probing window for one time point were then pasted into the column of a matrix and the procedure repeated over all time points to generate a Rac1 activity map ( Fig 7B ) . Next , we spectrally decomposed the protrusion activity map ( Fig 7C ) and performed SRM analysis on those spectral features of cell protrusion dynamics to identify distinct motion regimens ( Fig 7D ) . Using Q = 3 we found four distinct regions , each with a different average level of Rac1 activity ( Fig 7E ) . It should be noted that the motion regimens are transient in space and time . We visualized this behavior in a movie where the probing windows are color labeled in correspondence with their association to a particular motion regimen ( Video 9; Fig 7F displays selected snapshots at certain time points ) . In previous work[45] , we demonstrated that cycles of edge protrusion and retraction corresponded to cycles of Rac1 activity . Dependent on the distance from the edge , the motion and signaling cycles had distinct time lags and also showed different levels of correlation , which is a measure of their mutual association . While in the past we manually or semi-manually selected regions along the edge boundary suited for correlation analysis , we wondered whether the boundary regions identified by SRM would now in a more objective manner indicate differences in the magnitude and time lag of the correlation . We performed this region-based correlation analysis in a layer of probing windows at the edge and a second layer of windows shifted into the cell interior by 3 μm , thus covering a band 3–6 μm from the cell edge . Fig 7G–7J display the correlation functions for individual sectors ( blue ) and their average ( red ) for the four identified motion regimens in the first and second layers . Both regimens 1 and 2 displayed correlation functions with significant positive lobes for negative time lags and negative lobes for positive time lags in the first layer ( Fig 7G and 7H ) . Consistent with our previous analyses of Rac1 activation in protrusion-retraction cycles[45] , this meant that in these regions Rac1 activity was delayed by ~50–60 sec relative to cell protrusion , whereas Rac1 activity was minimal ~30–40 sec prior to protrusion events . Neither regimen 3 nor 4 displayed a significant correlation function in the first or second layer , indicating overall weaker Rac1 signaling in these regions , and especially a weaker coupling between edge motion and Rac1 activity ( Fig 7I and 7J ) . None of the four identified motion regimens displayed significant correlations between Rac1 activity in the second layer and edge motion . This is also consistent with our previous data[45] , which showed a decay of spatially finer sampled correlation values to insignificant values at 4 . 5 μm and longer distances . The correlation functions in the first layer for regimens 1 and 2 , however , showed a remarkable difference in the widths of the significant lobes . Regimen 1 had nearly symmetric lobes with a full width at half maximum ( FWHM ) of 40 sec , whereas regimen 2 had skewed lobes with a FWHM of 75 sec , We note that in previous analyses of correlations between molecular and cell protrusion activities such differences were obscured by the need for averaging over multiple sectors . It is tempting to speculate that the differences in signaling dynamics identified between regimen 1 and 2 are associated with different molecular programs driving Rac1 activity . With the presented SRM analysis of motion regimens , we now have the tool to systematically probe subcellular signaling activities that may even be transient in space and time , and relate them to cell morphogenesis and other cell functions .
In this work we implemented a framework for profiling cellular morphodynamics using spectral decomposition , instantaneous frequency analysis , and unsupervised clustering . First , we extracted the local dynamics of cell edge motion from time-lapse live cell image sequences by sampling protrusion and retraction velocities in discretized sectors of ~3 μm width along the cell periphery . Then , we conducted in every sector HHT-based spectral decomposition of the sampled velocity time series . The HHT resulted in several intrinsic mode functions , here fixed to six , each of which was transformed into instantaneous frequency and amplitude distributions . Hence , unlike a static spectral decomposition such as a Fourier Transform , the HHT-based decomposition captures variations in the oscillatory behavior between different time points and thereby allows detection of switches in the spectrum . A critical question to address in our development was how much the uncertainty of mapping the displacements of a cell edge between consecutive frames affects the spectral decomposition analysis . Given the previously published mapping algorithm ( see Methods ) , we performed a worst-case error assessment for the displacement data and then simulated how such an error level projects into the distribution of instantaneous frequencies ( S5 and S6 Figs ) . Specifically , our algorithm for computing edge displacements at the level of single pixel guarantees topological consistency among the virtual edge markers between consecutive frames , i . e . protrusion vectors are never allowed to cross each other ( S1 Fig ) . Hence , the maximal mapping error corresponds to the length difference Δdp of a vector that originates from a virtual edge marker at time t and targets one of the neighboring virtual marker positions at time t+1 ( S5A and S5B Fig ) . We then defined a relative mapping error rate ( S5B Fig ) , which for all marker points and an entire movie was approximately uniformly distributed between 0 and 8% , with an average mapping error rate of ~4% ( S5C Fig ) . To assess the effect of such errors on the spectral analysis we randomly perturbed an actual protrusion activity map ( S6A Fig ) with error rates spanning from 1% to 100% ( S6B–S6F Fig ) and computed the K-S statistics between the frequency distributions of original and perturbed protrusion time series ( S6G Fig ) . The simulations showed that mapping error rates of less than 10% generate deviations from the ground truth with K-S statistics less than the threshold value ~0 . 08 associated with the average variation of protrusion dynamics in an molecularly homogeneous cell population ( Fig 2F ) . Hence , we were assured that the 4% level of errors from the cell edge tracking per se does not significantly contribute to conclusions from spectral analysis of morphodynamic behaviors . Biologically , the most striking finding in this first study with HHT-based protrusion analysis is the distinct , nearly orthogonal meaning of instantaneous frequency and instantaneous amplitude spectra in terms of protrusion regulation . While the amplitude spectra report the speed of cell edge motion , the frequency spectra report how protrusion and retraction cycles are regulated . This is consistent with previous reports from our and other labs that have shown high sensitivity of measurements of protrusion persistence to perturbation of regulatory signals , whereas measurements of protrusion speed were largely unaffected by these same manipulations[44 , 46 , 58–60] . Cell protrusion requires on the one hand a process that stalls retraction and initiates forward edge motion . On the other hand , it requires a process that stabilizes and eventually reinforces the forward motion against increasing mechanical resistance by the environment and/or the stretched cell plasma membrane[42] . While initiation is stimulated by cell external signals or occurs spontaneously , as is the case for all data analyzed in this work , persistent edge advancement depends on the coordinated engagement of signaling pathways that converge on the activation of a series of nucleators and modulators for actin filament assembly after protrusion onset[61] . Many of these pathways are regulated by feedbacks , which integrate environmental and cell-intrinsic mechanical and chemical cues . Accordingly , dynamic or permanent changes in environmental cues , or in the pathways that process them , primarily affect the protrusion persistence . In contrast , the protrusion speed is less sensitive to the coordination of pathway engagement but more on the overall level of engagement . Moreover , the maximal velocity is reached in the early phases of protrusion , before the pathways critical for reinforcement are engaged[42 , 61] . This temporal separation of molecular processes that affect protrusion speed from processes that affect protrusion reinforcement explains also the orthogonality between maximal velocity and persistence measurements . To demonstrate the orthogonality of speed-related amplitude spectra and regulation-related frequency spectra we employed a recently developed optogenetic toolkit to instantaneously deactivate and reactivate a specific node in one of the regulatory pathways . We chose the GEF Vav2 , which is one of several activators of the Rac1 GTPase[38 , 56] implicated in the assembly of actin filaments required for lamellipodia-driven cell protrusion[62] . Given the redundancy of Vav2 with other Rac1 GEFs , we suspected that inactivation of Vav2 would cause a rewiring of the regulatory circuitry without completely shutting down the regulation process . Indeed , we found a light-dose dependent shift in the frequency spectra . The amplitude spectra , on the other hand , were remarkably stable across the range of applied Vav2 inhibition . This underscores the orthogonality between amplitude and frequency of spontaneous protrusion-retraction cycles , allowing us to distinguish molecular processes implicated in setting the activation level of the protrusion machinery from processes that control protrusion regulation . HHT-based profiles now introduce a refined framework to describe the state of the regulatory circuitry . Compared to the analysis of protrusion persistence , which requires time integration over multiple cycles , the decomposition of the motion signal into instantaneous frequencies allows distinction of spatially and temporally localized states of the circuitry . We therefore thought that the HHT-based spectra would allow us to define a multidimensional feature vector to distinguish edge sectors with transiently consistent morphodynamic behavior . We also hypothesized that these sectors would correspond to significantly different regulatory signaling activities , i . e . be associated with transient ‘signaling microdomains’[63] . We tested this by separately analyzing both the activation levels of Rac1 as well as the temporal correlation between Rac1 activity and cell edge motion in edge sectors belonging to distinct morphodynamic regimens . The temporal correlation is a surrogate for the coupling of Rac1 signaling with motion , i . e . how much the pathways downstream of Rac1 activation contribute to the morphodynamics analyzed by HHT-based profiling . Indeed , we found that cell edge regions with different morphodynamic behavior displayed different Rac1 activation patterns . This demonstrates that the profiling framework not only detects differences between cells with different molecular makeups , but also provides the means to identify subcellular regions with distinct signaling activities . A critical future application of this capacity will be in the analysis of signal transduction pathways implicated in the regulation of cell shape and migration . In the past , we have manually selected sectors obeying qualitative criteria of cell edge dynamics to perform image fluctuation-based analyses of signaling activities[44 , 45 , 47] , or have included all of the cell edge independent of their dynamics . In both cases , the spatiotemporal averaging necessary to extract meaningful information from signaling fluctuations was executed over the boundaries of signaling microdomains , resulting in less sensitivity and bias . Moving forward , HHT-based morphodynamic profiling and spatiotemporal clustering of similar profiles will be used to automatically identify sub-cellular regions of homogeneous signaling activities with high resolution in both time and space . A second future application of HHT-based frequency decomposition and spatiotemporal clustering will be in the time series analysis of activity biosensor fluctuations per se . While the presented work focused on domain definitions only of motion fluctuations , the same framework could also be applied to fluctuations in signal activation throughout the entire cells . This will potentially enable the complete mapping of sub-cellular signaling regimes , and in combination with perturbations of signaling nodes , the identification of sub-cellularly distributed functions of signals , which is currently experimentally inaccessible . In sum , HHT-based profiling and clustering will have numerous powerful applications in the quantitative analysis of cell behavior , from classifying whole-cell migration states to focusing on subcellular regulatory microdomains . The software to enable these types of analyses is accessible under the Github link: https://github . com/DanuserLab/MorphodynamicProfiling and https://github . com/DanuserLab/Windowing-Protrusion .
To identify the cell-edge location in the examples presented here , automatic thresholding was combined with morphological post-processing . Thresholds were automatically selected by fitting a smoothing spline to the image intensity histogram and by finding the first local minimum after the lowest-intensity maximum , thus selecting a threshold , which separates the low-intensity histogram mode corresponding to the background from the higher-intensity peak ( s ) associated with cellular fluorescence . In cases where this automatic approach failed , thresholds were manually selected . Images were pre-filtered with a Gaussian approximation of the point spread function prior to binarization by thresholding to minimize the effects of image noise . To ensure that the resulting segmentation contained only a single connected component corresponding to the cell , the thresholding was followed by automated morphological post-processing including hole-filling for small intracellular areas of low intensity , a closure operation to fill small gaps in low-intensity edge regions , and only the largest remaining connected component was retained to remove small background spots . Cell edge velocities were derived from pixel-by-pixel matching of cell contours between consecutive time points , as described in ref . [64] and reproduced here for completeness . In summary , a B-form spline was fitted to the edge pixel positions of the segmented cell area , with nodes corresponding to each edge pixel ( S1A Fig ) . The spline representations of two consecutive frames were then divided into segments between their intersections . To map a correspondence between the edge splines on consecutive frames , the following objective function was iteratively minimized: ( o^1 , … , o^n ) =argmin ( o^1 , … , o^n ) {∑i=1n[x ( t+1 , oi ) −x ( t , pi ) ]2+ω∑i=2n[oi ( t+1 ) −oi−1 ( t+1 ) pi ( t ) −pi−1 ( t ) ]2SUMASUMB} ( M . 1 ) withthetopologicalconstraintse1=o1<o2<…<on=en ( M . 2 ) The variable n denotes the number of nodes , which in the absence of down-sampling ( see below ) is equal to the number of edge pixels in that segment . p1 , 2 , …nt are the parameters of the spline at time t defining equally spaced edge nodes x ( t , pi ) , one at each edge pixel . The goal of Eqs M . 1 and M . 2 is to identify n spline parameters o1 , 2 , …nt+1 in between the intersection points e1 and en that define non-equally spaced nodes x ( t+1 , oi ) at t+1 such that the overall displacement ( SUMA ) and strain , i . e . changes in spacing of nodes ( SUMB ) is minimized . M . 2 imposes the constraint to the minimization that displacement vectors must not cross . The two sums in M . 1 have different physical units . To balance them correctly we introduce a factor ω as follows: ω=w* ( SUMASUMB ) iteration=1=w*{∑i=1n[x ( t+1 , oi ) −x ( t , pi ) ]2∑i=2n[oi ( t+1 ) −oi−1 ( t+1 ) pi ( t ) −pi−1 ( t ) ]2}iteration=1 ( M . 3 ) The factor ω is calculated only in the first iteration of the minimization , as the unit conversion by the ratio SUMA/SUMB changes insubstantially thereafter . The parameter w is a free user-control that allows the definition of the trade-off between minimal edge displacement and minimal lateral strain ( S1B–S1E Fig ) . For w = 1 these two competing criteria have equal weight . The global solution of the edge mapping is fairly insensitive to the value of w . However , adjustments may be useful to track particularly rugged features of the cell edge , or vice versa , to oppress the mapping of spiky edge features . The minimization of Eq M . 1 can be computationally costly when the number of edge pixels in a segment exceeds 100 . To circumvent this problem , we introduce a control parameter 10<Nmax<100 . When the number of edge pixels in a segment is greater than Nmax , we downsample the number of nodes to Nmax , calculate the boundary displacement for this number of nodes , and then up-sample to the original number of edge pixels by interpolation . This control parameter therefore not only allows the flexibility of trading computational speed for accuracy , but allows the method to be applied to cells of any size imaged at any resolution . Once the boundary displacements are identified , the projections of these displacements onto the boundary normal vector are calculated to obtain a signed local measurement of the instantaneous normal edge velocity . The nodes are reset with every time step ( S1F Fig ) . Accordingly , to compute a continuous path for a virtual edge marker throughout an entire movie it is necessary to interpolate marker positions for each time point ( not applied in this study ) . 0020 Our software supports two methods for sampling window creation . The first is a discrete pixel space method , which is faster and ensures windowing of the entire segmented area . The second is a sub-pixel method , which allows more flexibility and precision , but which excludes some segmented areas that do not meet strict criteria . The second method is also more computationally intensive . In both methods the intracellular frame of reference used to create the image sampling windows is based on the Euclidean distance transform D of the cell edge[65] . The discrete pixel space sampling window generation method creates sampling windows using both the discrete distance transform D and the nearest-neighbor transform or feature transform F: di=D ( ui , x1 , 2 , …n ) ( M . 4 ) fi=F ( ui , x1 , 2 , …n ) ( M . 5 ) where , for the ith pixel ui in the segmented cell , fi is the index of the closest pixel on the cell boundary x , and di is the distance to this pixel and therefore the shortest distance to the cell boundary . We then also calculate the associated distance along the cell boundary for each pixel , li=L ( ui ) =∑k=2fi|xk−xk−1| ( M . 6 ) The location of the origin of the sampling windows , x1 , is determined by the user . A given sampling window can then be defined as: Wm , p={u1 , 2 , … , I|ui∈Ω∧ bm<di≤bm+1∧ sp<li≤sp+1} ( M . 7 ) where Ω is the segmented cell area , b1 , 2 , …M are the user-selected distances from the cell edge , and s1 , 2 , …P are the user-selected distances along the cell edge . That is , a particular window Wm , p is defined as all pixels with distances between bm and bm+1 from the cell edge , and for which the distance from the origin along the closest cell edge is between sp and sp+1 . Note that in discrete pixel space it is non-trivial to define a distance measure L at a contour other than the cell boundary x . This is because with near convex cell edges the nearest feature transform , fi at positions inwards from the cell edge will not contain indexing which represents all of the pixels on the cell boundary . Therefore , the discrete windowing approach in our software package does not currently support specification of a ‘master contour’ other than the cell edge . In the sub-pixel windowing method , the cell interior is subdivided with respect to distance from the edge by defining isocontours ( or level sets ) C of the Euclidean distance transform D at distance isovalues specified by the user: Cm={u1 , ‥ , un|D ( u1 , ‥ , un ) =bm} ( M . 8 ) Where Cm is the mth isocontour at the distance value bm . Isocontour coordinates are refined to sub-pixel precision by linear interpolation of the original distance transform , which is calculated on the discrete pixel grid . The subdivisions of the cell interior into window slices are defined by first defining ‘slice’ start positions σ along a user-specified ‘master-contour’ Cμ: σ={Cμ , 1 , ‥ , Cμ , n|Lμ ( Cμ ) ∈s1 , 2 , …P} ( M . 9 ) Where s1 , 2 , …P are again the user-selected distances along the master contour μ and Lμ ( Cμ , i ) =∑k=2i|Cμ , k−Cμ , k−1| ( M . 10 ) is the distance along the master contour from the origin Cμ , 1 . The position of this origin can be set by the user as well , e . g . to mark the back or front of the cell . Note that this choice has no influence on the actual geometry of the sampling windows . The ‘slice’ curves S used to subdivide the cell from these slice start positions are then determined by a maximal-gradient ascent of the Euclidean distance transform: Sp , i=Sp , i−1+∇D ( Sp , i−1 ) ( M . 11 ) with Sp , 1=σp ( M . 12 ) and the local gradients are again estimated via linear interpolation . The geometry of an individual sampling window Wm , p is then defined as the area enclosed by the two isocontours Cm and Cm+1 and the slice curves Sp and Sp+1 . This ensures that the image area within each sampling window occupies a specific range of distances from the cell edge , and that the closest region of the cell edge is the one delineated by the intersection of gradient ascent polygons and the cell edge . Regions of the cell interior that do not meet these criteria are excluded from the windowing . This includes regions spanning ridges in the distance transform , which are therefore equally proximal to two disconnected regions of the cell edge , and regions near image borders , where the association with the cell edge is indeterminate . In an image time-lapse sequence , the position of the sampling windows in each frame can be determined in several ways: The algorithm described above can be applied to each frame using constant isocontour and gradient ascent curve spacing , and only the location of the origin varies with time . The location of this origin is propagated between subsequent frames either by using the closest edge displacement vector or by finding the closest point on each subsequent cell edge to the original user-selected origin . In either of these cases the number of sampling window slices can vary with respect to time if the length of the cell edge changes . Alternatively , the number of sampling window slices can be held constant , allowing the width of each to vary as the length of the cell edge changes . Finally , it is also possible to allow each gradient ascent start-point to follow the edge-displacement vectors adjacent to it . This propagation method will maintain the number of window sampling slices , but will allow each slice to expand or contract as the associated region of the cell edge protrudes or retracts . This last method tends to generate the most stable window configurations . Irrespective of the propagation method chosen , each sampling window band will always maintain its distance from the cell edge . For all analyses in Fig 7 we used the setting with constant number of window sampling slices . Once the sampling windows are generated for each image of a dataset , the associated image signals , image-derived data and edge velocities can be sampled . For image sampling , a variety of statistics are calculated ( mean , standard deviation , maximum etc . ) for each pixel whose center lies within a given sampling window , yielding sample statistics in the activity matrix for each sampling window at each time point . For sampling of edge velocities , statistics are calculated for the displacement vectors associated with a particular cell edge sector . For example , instantaneous cell edge velocities are calculated for each edge sector as the average of the projections of the displacement vectors onto the associated edge normal divided by the time interval between frames . Because the sample sizes per edge sector may vary with the local cell edge geometry and motion , the number of edge displacement vectors contributing to each sample is also quantified . Cos7 cells were maintained in DMEM growth medium supplemented with 10% ( vol/vol ) FBS at 37°C and 10% CO2 . The transient transfection of Cos7 cells were performed using Fugene 6 transfection reagent ( Promega ) under the guidelines of the manufacturer . The YFP-PI ( WT ) -Vav2 ( DPZ ) ( Addgene #86974 ) and mCherry-lifeAct expressing plasmids were used to photo-inhibit Vav2 and label actin[38] . For 48 spontaneously migrating cells , a plasmid expressing mCherry-lifeAct was used . To monitor Rac1 activity , cells were transfected with dual chain Rac1 biosensor Rac1FLARE . dc1g[45] that has a dTurquoise and YPet fluorescent protein pair . To obtain a fixed ratio of two chains , two consecutive 2A viral peptide sequences from porcine teschovirus-1 ( P2A ) and Thosea asigna virus ( T2A ) were inserted between two chains , leading to cleavage of the two chains during translation . For live cell imaging , cells were plated on sterilized coverslips coated with 5 μg/mL of fibronectin ( Sigma ) and incubated in DMEM growth medium supplemented with 10% ( vol/vol ) FBS at 37°C . On the day of imaging , cell medium was replaced with L15 imaging medium ( Invitrogen ) supplemented with 5% ( vol/vol ) FBS . The coverslips with cells were placed in an open heated chamber ( Warner Instruments ) and live cell imaging was performed with an Olympus IX-81 inverted epifluorescence microscope equipped with an Olympus 40x UPlan FLN1 . 25 N/A silicon oil objective and a Flash 4 sCMOS camera ( Hamamatsu ) with temperature control ( BC-100 20/20 Technology ) . For excitation , a 100 Watt mercury arc lamp with a 3% ND filter and a 510–520 ( YFP ) nm or 565–595 ( mCherry ) nm band-pass filter was employed . A 1% ND filter and a 426–446 nm band-pass filter were used with a 100 Watt mercury arc lamp ( ~ 1 nW/μm2 of power density at λ = 488 nm , measured at the specimen plane ) for blue light pulse illumination . For emission ratio imaging of Rac1 , CFP: ( ex ) FF-434/17 , ( em ) FF-482/35; FRET: ( ex ) FF-434/17 , ( em ) FF-550/49; YFP: ( ex ) FF-510/10 , ( em ) FF-550/49 filters ( Semrock ) were used . Images of control cells and cells expressing the PI-Vav2 , and cells expressing Rac1 biosensor were taken every 10 and 5 sec , respectively . | Many studies in cell biology employ global shape descriptors to probe mechanisms of cell morphogenesis . Here , we implement a framework in this paper to profile cellular morphodynamics very locally . We employ the Hilbert-Huang transform ( HHT ) to extract along the entire cell edge spectra of instantaneous edge motion frequency and magnitude and use them to classify overall cell behavior as well as subcellular edge sectors of distinct dynamics . We find in fibroblast-like COS7 cells that the marked heterogeneity in mobility of an unstimulated population is fully captured by differences in the magnitude spectra , while the frequency spectra are conserved between cells . Using optogenetics to acutely inhibit morphogenetic signaling pathways we find that these molecular shifts are reflected by changes in the frequency spectra but not in the magnitude spectra . After clustering cell edge sectors with distinct morphodynamics we observe in cells expressing a Rac1 activity biosensor that the sectors with different frequency spectra associate with different signaling intensity and dynamics . Together , these observations let us conclude that the frequency spectrum encodes the wiring of the molecular circuitry that regulates edge movements , whereas the magnitude captures the activation level of the circuitry . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
]
| [
"velocity",
"biosensors",
"cell",
"physiology",
"cell",
"motility",
"classical",
"mechanics",
"engineering",
"and",
"technology",
"light",
"electromagnetic",
"radiation",
"cell",
"polarity",
"developmental",
"biology",
"mathematics",
"statistical",
"distributions",
"light",
"pulses",
"probability",
"theory",
"physics",
"detectors",
"cell",
"biology",
"equipment",
"cell",
"migration",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"motion"
]
| 2018 | Profiling cellular morphodynamics by spatiotemporal spectrum decomposition |
During cytokinesis , a contractile ring generates the constricting force to divide a cell into two daughters . This ring is composed of filamentous actin and the motor protein myosin , along with additional structural and regulatory proteins , including anillin . Anillin is a required scaffold protein that links the actomyosin ring to membrane and its organizer , RhoA . However , the molecular basis for timely action of anillin at cytokinesis remains obscure . Here , we find that phosphorylation regulates efficient recruitment of human anillin to the equatorial membrane . Anillin is highly phosphorylated in mitosis , and is a substrate for mitotic kinases . We surveyed function of 46 residues on anillin previously found to be phosphorylated in human cells to identify those required for cytokinesis . Among these sites , we identified S635 as a key site mediating cytokinesis . Preventing S635 phosphorylation adjacent to the AH domain disrupts anillin concentration at the equatorial cortex at anaphase , whereas a phosphomimetic mutant , S635D , partially restores this localization . Time-lapse videomicroscopy reveals impaired recruitment of S635A anillin to equatorial membrane and a transient unstable furrow followed by ultimate failure in cytokinesis . A phosphospecific antibody confirms phosphorylation at S635 in late cytokinesis , although it does not detect phosphorylation in early cytokinesis , possibly due to adjacent Y634 phosphorylation . Together , these findings reveal that anillin recruitment to the equatorial cortex at anaphase onset is enhanced by phosphorylation and promotes successful cytokinesis .
In cytokinesis , cells assemble and stabilize an actomyosin ring between segregated chromosomes to generate daughter cells . The positioning of the cleavage furrow is controlled by negative signals from astral microtubules and positive signals established from the central spindle [1 , 2] . At anaphase onset , inactivation of cyclin-dependent kinase 1 ( Cdk1 ) triggers recruitment of centralspindlin to the central spindle and adjacent equatorial cell membrane [3–5] . Centralspindlin recruits the RhoGEF , epithelial cell transforming sequence 2 ( Ect2 ) , to locally activate the small GTPase RhoA [2 , 4 , 6] which specifies localization of the contractile actomyosin ring [7] . Temporal recruitment and activation of the centralspindlin apparatus depends on phosphorylation , including Cdk1-dependent phosphorylation sites on mitotic kinesin-like protein 1 ( MKLP1 ) and Ect2 , lost at anaphase onset [3 , 6] , and anaphase-specific recruitment of Plk1 through phosphorylation of protein regulating cytokinesis 1 ( PRC1 ) [8] . Phosphorylation is likely to regulate additional events in cytokinesis . Anillin is a key scaffold protein linking the actomyosin ring to the equatorial membrane [9–11] . Anillin binds myosin and F-actin at the N-terminus , and has anillin homology ( AH ) and pleckstrin homology ( PH ) domains at the C-terminus ( Fig 1A ) . Anillin’s myosin and F-actin binding domain are required for organization of the actomyosin ring [12 , 13] . The AH domain of anillin binds RhoA and shares homology with Rhotekin , a RhoA-GTP binding domain [10] . Drosophila anillin also binds to RacGAP50C , a homologue of MgcRacGAP , suggesting that it may couple the central spindle to a contractile ring [14 , 15] . The C-terminal PH domain is important for recruitment of anillin to the equatorial membrane . Recently , a cryptic C2 domain within the AH domain was discovered and , with adjacent PH domain , promotes efficient recruitment to the membrane [16] . Thus , anillin is a hub for midzone membrane regulators and effectors of cytokinesis . Highlighting its required role in these processes , anillin depletion results in cytokinesis failure [13 , 17 , 18] . Despite its importance , poorly defined mechanisms restrict anillin to operate specifically at cytokinesis and at the equatorial membrane . Anillin localization varies with the cell cycle . In interphase , anillin primarily localizes to the nucleus [9 , 12] , but upon entry into mitosis it re-localizes uniformly to the cell cortex . At anaphase onset , anillin is lost from the poles and concentrates at the equatorial zone , prior to onset of cytokinesis [19] . The loss adjacent to the poles is explained in part by Ran-GTP signals which emanate from chromatin [20] . One candidate for anillin recruitment is concentration of its preferred lipid phosphatidylinositol 4 , 5-bisphosphate ( PI ( 4 , 5 ) P2 ) at the equatorial membrane [21] . Locally activated RhoA promotes anillin recruitment to the equatorial membrane and anillin stabilizes RhoA against extraction with trichloroacetic acid ( TCA ) [10 , 18 , 19] . Although these may be sufficient to specify timely recruitment of anillin , additional regulatory mechanisms of temporal-spatial control could reinforce these signals . Protein phosphorylation is a common mechanism of regulating localization and function in mitosis . Anillin displays phosphorylation-induced band retardation in SDS-PAGE [22] and large-scale phosphoproteomics identified numerous phosphorylation sites [23 , 24] . In fission yeast , the Polo-like kinase Plo1 binds to and phosphorylates an anillin-like protein Mid1 , facilitating contractile ring assembly [25] . Similarly , phosphorylation sites of S . cerevisiae anillin homolog , Bud4 , by Cdk1 promotes cytokinesis [26] . However , human anillin is divergent from its yeast counterparts , and several phosphorylation sites are not conserved , suggesting that the regulation may likewise differ . We hypothesized that phosphorylation of anillin governs its required function in human cytokinesis . Here we report that human anillin is indeed phosphorylated in mitosis , and that phosphorylation is required for proper cytokinesis . To evaluate functional significance of phosphorylation , we used a stringent assay to comprehensively evaluate non-redundant phosphorylation events on anillin . Surprisingly , most phosphorylation events are dispensable for cytokinesis . However , we identify phosphorylation of a single residue , S635 as a key determinant of anillin localization and function . This phosphorylation site controls its ability to efficiently concentrate on the equatorial plasma membrane , and maintain a stable cytokinetic furrow . This finding suggests that phosphorylation at this site controls timely equatorial accumulation of anillin required for cytokinesis .
To investigate anillin phosphorylation we first characterized how it is modulated in human mitosis . In mitotic HeLa extracts , anillin migrated slowly on SDS-PAGE compared to its S-phase counterpart ( Fig 1B ) . This shift was reversed by Lambda phosphatase , indicating the retardation is attributable to phosphorylation . Consistent with phosphorylation being specific to mitosis , the slow-migrating band diminished upon mitotic exit ( Fig 1B , right lanes ) . These observations were not specific to HeLa; similar mitotic shifts were observed for anillin in other cancer cells and non-transformed human cells ( Fig 1C ) . We conclude that anillin is phosphorylated significantly in human mitosis . To identify kinases that might be responsible for mitotic phosphorylation of anillin , we purified recombinant GST-tagged fragments and performed in vitro kinase assays with the mitotic kinases polo-like kinase 1 ( Plk1 ) , Aurora B , and Cdk1 . Under these conditions , all phosphorylated anillin , though on different domains: Plk1 and Cdk1 phosphorylated the 371–607 fragment , whereas Aurora B preferred the PH domain ( Fig 1D ) . To test if one kinase is primarily responsible for the electrophoretic mobility shift , we employed specific inhibitors of Plk1 , Aurora B , and Cdk1 . Inhibition of Cdk1 with RO-3306 restored fast electrophoretic mobility not seen with other inhibitors ( S1 Fig ) . Thus phosphorylation events that shift anillin mobility are governed directly or indirectly by Cdk1 , although these findings do not exclude possible roles by Plk1 or Aurora B , as some phosphorylation events might not significantly alter shift . Additionally , other kinases could operate on anillin , such as Citron or Rho kinases , which have a known role in cytokinesis [27] . We conclude that anillin is phosphorylated in mitosis and is regulated by mitotic kinases , consistent with a possible regulatory role in cytokinesis . As multiple mitotic kinases can phosphorylate anillin , we focused on residues known to be phosphorylated without kinase bias . From two phosphorylation site databases , we identified 46 sites , the majority discovered by phosphoproteomic mass spectrometry [23 , 24] . These 46 sites ( Ser , Thr or Tyr ) were fragmented into 7 subdomains , and we made constructs for each , replacing all with non-phosphorylatable residues ( Ala , Val or Phe , respectively ) ( Fig 2A ) . For the mutants containing a large number of phosphorylation sites ( A1-A5 ) , we employed the Gibson assembly method to assemble blocks of synthetic double-stranded DNA [28] . To study the functions of these phosphorylation events , we optimized an RNAi knockdown and transgene rescue assay ( Fig 2B ) . This assay was developed under conditions to establish the most robust difference between the GFP and WT controls to allow functional assessment of anillin mutants . In an asynchronous population of cells fixed after 48 h of RNAi treatment , 79% of GFP-positive cells were multinucleate . This phenotype was efficiently suppressed by coexpression of siRNA-resistant wild-type anillin ( Fig 2C and 2D ) . Under these conditions , the GFP-tagged constructs in panel A were assessed for their ability to restore cytokinesis . Strikingly , most non-phosphorylatable constructs suppressed multinucleated cells similar to wild type . However , the A5 mutant had a two-fold increase in multinucleation , despite efficient expression . To evaluate the efficacy of knockdown/addback under these conditions , we performed quantitative immunoblotting with infrared secondary antibodies ( Fig 2E ) , revealing that knockdown was ~80% effective , and that transfection added back anillin below endogenous levels . We conclude that most phosphorylation sites are dispensable in this assay , but that one or more of these sites in the g5 domain play a crucial non-redundant role in successful cytokinesis . Because anillin predominantly localizes at the cleavage furrow in anaphase , we tested if the A5 mutant impairs localization to the furrow . To do this , HeLa cells were transiently transfected with either GFP-tagged wild-type or A5 anillin , enriched at anaphase and assessed for localization . Endogenous anillin was used as internal control to verify that each cell is evaluated at the time proper for anaphase localization . Indeed , we found that A5 anillin remains diffused throughout the cytoplasm in early and late anaphase , wherein endogenous anillin concentrates at furrow ( Fig 3A ) . This phenotype was scored in two independent ways: a cell population scored by a blinded observer where only cells with strong furrow localization was counted ( Fig 3B ) , and an intensity analysis in which the GFP fluorescence at the equatorial membrane was normalized to the cytoplasmic levels near poles in a single cell ( Fig 3C ) . Both analyses confirmed impaired recruitment of A5 anillin at the furrow . We conclude that the A5 construct , with 11 nonphosphorylatable mutations , is not efficiently recruited to the equatorial midzone on anaphase onset . RhoA is a central regulator of furrow formation [29 , 30] and its TCA-fixable cortical pool is dramatically reduced by loss of anillin [10] . To test if A5 was able to stabilize active RhoA , cells were fixed/extracted with TCA [31] . As expected , fixed cortical RhoA is significantly reduced at both early and late anaphase cells expressing A5 anillin ( Fig 3D and 3E ) . These findings suggest that A5 anillin has impaired equatorial recruitment and is unable to fix cortical RhoA at the furrow . Phosphorylation sites could operate distributively to enhance furrow recruitment of anillin . Alternatively , a single phosphorylation site could be primarily responsible . To test this , we generated single/double non-phosphorylatable mutants for the each of the 11 sites in A5 anillin . Each GFP-tagged mutant was transiently expressed in HeLa cells and analyzed for localization . Most single/double mutants localize properly like their endogenous counterparts ( S2 Fig ) . However , anillin Y634F/S635A behaved like A5 , showing a dispersed localization ( Fig 4A ) . To distinguish the relative contributions of Y634 and S635 , we tested mutations singly . Both mutants disrupt anillin localization , but S635A had a more marked contribution ( Fig 4B–4D ) . Thus , from 11 sites , we identified S635 as a major residue required for proper anaphase localization of anillin . Moreover , S635 is conserved across metazoans , supporting its role as a critical phosphorylation site for cytokinesis ( Fig 4E ) . In principle , loss of the hydroxyl group in S635A anillin could disrupt protein function by a mechanism other than preventing phosphorylation . If phosphorylation per se is important , its function could be restored by a phosphomimetic negatively-charged residue at this site . Indeed , anillin S635D concentrated at the equatorial cortex ( Fig 4F ) ; this was clearly distinguishable from the one of S635A , as judged by a blind observer assay ( Fig 4G ) or by quantitative immunofluorescence ( Fig 4H ) , although it did not restore recruitment to the furrow to the extent of wild type . This disrupted localization is specific to anaphase , as anillin S635A localized properly in interphase and metaphase ( S3 Fig ) . Thus , S635 is crucial for anillin localization at the furrow but dispensable for localization at other times . These results support the idea that phosphorylation of S635 is critical for anillin recruitment to equatorial membrane in cytokinesis . To characterize the cytokinesis defects and membrane recruitment , we performed timelapse videomicroscopy of wild type and mutant anillin constructs after depleting endogenous anillin ( Fig 5 and S1–S5 Videos ) . With depletion of anillin and GFP transfection , we observed marked furrow instability , as expected [10 , 13 , 19] . This effect was rescued with GFP-wildtype anillin , which was efficiently recruited to membrane and enriched at the furrow by 5–9 minutes after anaphase onset . By contrast , both S635A and double mutant Y634F/S635A were poorly recruited to membrane during mitosis , although some furrow localization is seen . These constructs failed to restore furrow stability and resulted in significant oscillation of furrows . The phosphomimetic , S635D , restores membrane recruitment , enrichment at the cleavage furrow , and stabilization of the mitotic furrow . To assess this in a larger population of cells , we analyzed furrows in timelapse videomicroscopy of HeLa cells expressing mCherry-H2B ( S4 Fig ) . As above , these images reveal transient furrows and then failed cytokinesis to generate a single cell with two nuclei . To assess frequency of furrow instability with non-phosphorylatable anillin , we developed a fixed-cell assay to identify cells with unstable furrows ( Fig 6A and 6B ) . As expected anillin depletion increased the number of cells with eccentric furrows . Likewise , non-phosphorylatable mutants had increased eccentric furrows compared with wild-type control . Thus , the non-phosphorylatable mutants of anillin are unable to sustain a stable contractile ring , consistent with poor recruitment of these mutants in early anaphase . Together , these data provide high confidence that S635 is important to sustain the mitotic furrow . We next considered mechanism for the cortex association of phospho-S635 anillin . The AH domain of anillin adjacent to S635 is known to interact with RhoA and Ect2 [10 , 32] . Therefore , we reasoned that phopho-S635 anillin might relieve autoinhibition of the AH domain to stabilize its interaction with these interactors . To address this , we performed a pulldown assay from mitotic HeLa extracts with either unphosphorylated ( GST WT AH-PH ) or the phosphomimetic ( GST S635D AH-PH ) C-terminal anillin fragment . For RhoA pull down , cells were transfected with constitutively active RhoA ( Q63L ) . However , RhoA and Ect2 were pulled down equally well by the wild-type and phosphomimetic C-terminal anillin fragment ( Fig 6C ) . We conclude that S635 phosphorylation is unlikely to regulate interaction with Ect2 or active RhoA . A key localization requirement for anillin is phosphatidylinositol phosphate at the plasma membrane , mediated by its C-terminal PH domain [10] . This prompted us to test the effect of phosphorylation at S635 on association with membrane phospholipids . GST-tagged anillin AH-PH variants were incubated with an array of lipids that were immobilized on a support membrane . GST fusions of both wild-type and phosphomimetic AH-PH bound to phosphatidylinositol 4-phosphate PI ( 4 ) P , PI ( 4 , 5 ) P2 , and PI ( 3 , 4 , 5 ) P3 . However , there was no observed difference in preference for phospholipids ( Fig 6D ) . Thus the specific mechanism by which phosphorylation regulates anillin recruitment and furrow stability appears to be distinct from its roles in binding RhoA , Ect2 , and does not control specificity for binding of the C-terminus to specific phospholipids . To evaluate localization of phosphorylation , we raised a polyclonal antibody against a peptide encompassing phospho-S635 , and validated it by several measures . First , dot-blot verified >1000 fold phospho-specificity , as non-phosphorylated peptide was not detected at up to 1250 pmol ( Fig 7A ) . By Western blot of whole cell lysates , pS635 antibody recognized bands from HeLa lysates that matched those detected by anti-anillin antibody and decreased upon anillin depletion ( S5 Fig ) . We were unable to fully validate specificity for pS635 by immunoblot . By contrast , immunofluorescence confirmed reactivity of the phosphospecific antibody at the midbody in late cytokinesis , whereas the non-phosphospecific antibody did not detect an epitope at this site ( Fig 7B ) . Finally , siRNA against anillin/addback revealed that the phosphoepitope is detected with wild-type anillin transfection , but not with S635A ( Fig 7C and 7D ) . We conclude that this antibody specifically detects phosphorylated S635 by immunofluorescence . Having established the specificity of the pS635 antibody in immunofluorescence , this reagent was used to investigate the dynamics of phosphorylation during cytokinesis . Localized S635 phosphorylation appears in cells after furrowing and persists until late cytokinesis ( Fig 7E ) . Strongly enriched signal is seen at the cortical midzone at late cytokinesis . We performed a more detailed analysis of pS635 signal at different stages of anaphase under different fixation/extraction conditions and confirmed that the signal is only seen in late mitosis ( S6 Fig ) . However , it was puzzling that pS635 is not detected earlier in anaphase when anillin is being recruited to membrane . We considered that the adjacent post-translational modifications could partially interfere with the antibody . Indeed , the antibody fails to detect a doubly phosphorylated peptide at both Y635 and S635 ( S7 Fig ) . We conclude that S635 on anillin is an in vivo phosphorylation site , although our reagent cannot detect early mitotic phosphorylation , possibly due to interference with adjacent post-translational modifications .
Anillin is phosphorylated in mitosis [22] , but heretofore , the physiological role of these post-translational modifications has been obscure . We discovered that among 46 phosphorylated residues , S635 is required for efficient anillin recruitment to the furrow and for successful cytokinesis . Several lines of evidence confirm that S635 is important due to phosphorylation at this site rather than to another function of serine hydroxyl . First , it can be detected with a phospho-specific antibody on anillin in late cytokinesis here and previously with mass spectrometry . Second , its localization outside mitosis is unaffected by this mutation . Third , its function is partially restored with a phosphomimetic aspartic acid residue , demonstrating that negative charge at this site is important for its function , but the serine hydroxyl is dispensable . Finally , this residue is phylogenetically conserved in metazoans . Together , these data strongly support a role of phosphorylation at this site to allow anillin to function in cytokinesis . The kinase responsible for S635 phosphorylation remains obscure . The amino acid sequence surrounding S635 does not match canonical motifs of mitotic kinases . Moreover , in preliminary experiments , we did not observe phosphorylation in the domain containing S635 by Cdk1 , Plk1 , or Aurora B . However , it is possible that cell contexts or modification of adjacent residues ( such as Y634 phosphorylation ) could alter kinase specificity . To identify the kinase and understand phospho-regulation of anillin , it will be important to consider a broad host of kinases involved in mitosis , especially those required for cytokinesis , including Rho-associated kinases . We were unable to identify a specific mechanism by which S635 phosphorylation regulates recruitment of anillin to the equatorial membrane in anaphase . The AH-PH domain is sufficient for membrane recruitment without the region encompassing S635 [10 , 16] . However , the fragment encompassing S635 could be an autoinhibitory domain , precluding AH-PH domain binding to membrane until anaphase . Phosphorylation of S635 could relieve this autoinhibitory control of AH-PH to regulate timely membrane association in anaphase . Mechanistic and structural experiments are needed to determine if this phosphorylation controls affinity of the anillin N-terminus for efficient membrane recruitment . Moreover , it will be important to assess how phosphorylation controls membrane interaction . One paradoxical finding is that pS635 anillin is observed primarily late in cytokinesis , yet our data suggest phosphorylation at this site is required for early anillin recruitment . One possibility is that phosphorylation of anillin occurs only late in anaphase and controls its roles in abscission [33–35] . This model is consistent with the observation of late cytokinesis failure but appears inconsistent with the impaired recruitment and the furrow instability seen with S635A . Another possible explanation for failure to detect pS635-anillin in early anaphase is that the phospho-antibody has less affinity for pS635-anillin than the total anillin antibody , so the pS635-anillin is only visualized after the ingressing furrow concentrates it in late cytokinesis . A final explanation is that phosphorylation at Y634 masks the pS635 in early cytokinesis , as we observe with phosphopeptide analysis . Thus it is possible that phosphorylation at this site occurs early in mitosis and impairs recognition until late in cytokinesis . In any case , the antibody data provide further confidence that anillin is phosphorylated at this site in human cells . Although it may seem remarkable that many anillin phosphorylation sites appear dispensable for cytokinesis , our data do not allow for this conclusion . First , our depletion was optimized for detecting/rescuing binucleation and resulted in only ~80% knockdown of endogenous anillin; it is possible that phosphorylation on residual anillin was sufficient to retain some functions of anillin . Second , the transfection assays are heterogeneous , and it is possible that high expression in some cells masked defects of non-phosphorylatable mutants . Third , we evaluated phosphorylation sites by domain-specific mutants . Although many of the domain-specific mutants , A1-4 , and A6-7 , appeared to function for cytokinesis , it is possible that redundant phosphorylation sites span the boundaries we selected or , perhaps , our assay was not sensitive to subtle functions within cytokinesis . Moreover , we find that mitotic phosphorylation is decreased when Cdk1 is inactivated , possibly removing a negative regulator of anillin function in early mitosis . If so , there are additional layers of regulation as mutations of multiple putative Cdk1 phosphorylation sites did not ultimately preclude successful mitosis and cytokinesis . Additionally , phosphorylation at other sites may be important for differentially regulated functions of anillin , such as meiotic division [33] . Previously known regulatory mechanisms are insufficient to wholly explain the temporal and spatial control of anillin in cytokinesis . For example , astral microtubules [36] and Ran-GTP signals from chromatin [20] , can help exclude anillin from polar membrane , although the latter appears to require close apposition of chromatin . Additionally , recruitment of anillin and RhoA to the membrane are mutually dependent [10 , 18 , 19] , suggesting RhoA activation could be partly responsible for timing . Temporal and spatial control of the upstream kinase and anillin phosphorylation can further enhance the specificity of anillin recruitment . In sum , using a functional screen , we identify an essential phosphorylation at S635 , important to reinforce the equatorial zone of anillin . This phosphorylation site provides temporal control of its interaction with the equatorial membrane . This allows anillin to efficiently integrate Rho with its upstream regulators and downstream regulators in a timely fashion to ensure successful cytokinesis . It will be important to identify the kinase responsible for this phosphorylation and to understand how it operates with convergent mechanisms for timely and specific recruitment of anillin for cytokinesis .
Full-length human anillin cDNA isoform 2 ( accession number BC070066 , Open Biosystems ) was cloned into pEGFP-C1 ( Clontech ) . This isoform was used previously to analyze anillin domain functions and encodes a protein with a 37 residue gap between the actin-binding and rho-binding domains , compared with the longest isoform . An RNAi-resistant version was made by polymerase chain reaction ( PCR ) , by engineering of the following changes: nt 798 5’-TGCCTCTTTGAATAAA-3’ 814 to 5’-CGCAAGCTTAAACAAG-3’ , creating silent mutations in the cDNA . Various derivatives were made from the RNAi-resistant template for rescue experiments . GST fusions were generated by subcloning of various anillin fragments into the pGEX-6P-1 vector ( GE Healthcare ) . The phosphodeficient subdomains of anillin ( gBlocks: g1-g5 ) were generated as double-stranded , sequence-verified genomic blocks by Integrated DNA Technologies ( IDT ) . gBlock fragments and PCR amplified anillin fragments were added to Gibson Assembly Master Mix ( New England Biolabs ) and incubated at 50°C for 1 h . Assembled full-length phosphodeficient mutants of anillin are then cloned into pEGFP-C1 . Cell lines were propagated at 37°C in 5% CO2 in media supplemented with 10% fetal bovine serum and 100 units/ml penicillin-streptomycin . The following media were used: DMEM ( HeLa and ACHN ) , DMEM:F12 ( RPE ) , RPMI ( 786-O ) . To enrich cells in anaphase , cells were treated with monastrol for 8 h , and fixed 60 min after release from the monastrol block . The following siRNA duplexes were used: control ( Thermo Scientific siGENOME Non-Targeting siRNA #2 D-001206-14 ) , anillin ( 40 nM; Thermo Scientific , custom order; CGAUGCCUCUUUGAAUAAA ) . Lipofectamine 2000 ( Invitrogen ) was used for siRNA/add back transfection . Cells were analyzed 48 h after transfection . For transient DNA transfection , HeLa cells were transfected using FuGENE HD ( Promega ) and analyzed 24 h to 48 h post transfection . GST-tagged anillin fragments were expressed in E . coli ( BL21 ) and expression was induced by the addition of 0 . 1 mM IPTG at 30°C for 4–5 h . Bacteria were resuspended in PBS containing 250 mM NaCl , 10 mM EGTA , 10 mM EDTA , 0 . 1% Tween-20 , 1 mM DTT , 1 mM phenylmethanesulfonylfluoride ( PMSF ) and 1 mg/ml lysozyme prior to sonication . Lysates were purified using Glutathione Sepharose 4 Fast Flow ( GE Healthcare ) . For the in vitro kinase assay , a series of GST-tagged anillin fragments ( 2 μg ) were incubated with each kinase in Mg+2-containing kinase buffer with 0 . 1 mM ATP and 1 μCi [γ-32P] ATP at 30°C for 30 min . The reaction was terminated by addition of SDS sample buffer . Samples were resolved by SDS-PAGE , visualized by Coomassie brilliant blue staining , and finally analyzed using a phosphor imager ( Typhoon , GE Healthcare ) . HeLa cells were lysed in lysis buffer ( 50 mM HEPES pH 7 . 5 , 100 mM NaCl , 0 . 5% NP-40 , 10% glycerol , 1 mM DTT , 1 mM PMSF , protease inhibitor cocktail and phosphatase inhibitor cocktail ) and incubated with purified GST-tagged anillin AH-PH at 4°C for 4 h . The beads were washed twice with lysis buffer and subjected to SDS-PAGE . Pulled down proteins were detected by immunoblotting using anti-RhoA or anti-Ect2 antibody . For polyclonal anillin antibodies , GST-tagged anillin amino acids 372–607 were expressed in E . coli , purified by Glutathione Sepharose 4B ( GE Healthcare ) . Tag-cleaved proteins by Prescission protease were then used to immunize rabbits for production of antisera . Raw serum was tested for the specificity of antibodies by Western blot analysis , immunoprecipitation , and immunofluorescence staining ( S5 Fig ) . To make phosphospecific antibodies , phosphopeptides were used to immunize rabbits . Serum was passed through a phosphopeptide affinity column . The eluted antibodies that contain a mixture of phospho- and nonphospho- antibodies were passed through a nonphosphopeptide column ( Genemed Synthesis Inc . ) . The flow through was tested for phosphospecificity . The following primary antibodies were used: rabbit anti-anillin ( 1:2000; IF , 1:1000; blot ) , mouse monoclonal anti-GFP ( 1:1000; 3E6 , Invitrogen ) , mouse monoclonal anti-cyclin B1 ( 1:2000; BD Biosciences ) , mouse monoclonal anti-β actin ( 1:20 , 000 , Abcam ) , mouse monoclonal anti-RhoA ( 1:1000; 26C4 , Santa Cruz Biotechnology ) , rabbit anti-Ect2 ( 1:1000 , Santa Cruz Biotechnology ) , mouse monoclonal anti-GST ( 1:2000; B-14 , Santa Cruz Biotechnology ) , mouse monoclonal anti-Flag ( 1:2000; M2 , Sigma-Aldrich ) , rat monoclonal α-tubulin ( 1:1000; YL1/2 , Millipore ) . Reagents used in this study are Lambda protein phosphatase ( New England Biolabs ) , Gibson Assembly Master Mix ( New England Biolabs ) , nocodazole ( 0 . 2 μg/ml; EMD Biosciences ) , monastrol ( 100 μM; R&D Systems ) , thymidine ( 2 . 5 mM; EMD Biosciences ) , BI-2536 ( 200 nM; a gift from P . Jallepalli ) , ZM-447439 ( 4 μM; R&D Systems ) , RO-3306 ( 10 μM; R&D Systems ) . For immunofluorescence ( IF ) , cells were cultured on glass coverslips in 24-well plates and fixed with 4% paraformaldehyde or ice-cold TCA for 15 min . Fixed cells were then blocked for 30 min in 3% bovine serum albumin ( BSA ) and 0 . 1% Triton X-100 in PBS ( PBSTx + BSA ) . Primary antibodies were incubated in PBSTx + BSA for 1 h at room temperature and washed three times in PBSTx followed by secondary antibody incubation in PBSTx + BSA for 30 min at room temperature and two washes with PBSTx . Cells were counterstained with DAPI , mounted on glass slides with Prolong Gold antifade medium ( Invitrogen ) , and allowed to cure overnight . Image acquisition was performed on a Nikon Eclipse Ti inverted microscope equipped with CoolSNAP HQ2 charge-coupled device camera ( Photometrics ) . To visualize GFP-tagged anillin and high resolution imaging of cytokinesis furrow formation , HeLa cells were grown in 12-well plates and transiently transfected with siRNA and anillin cDNA . 24 hr after transfection , cells were transferred to 1 . 5 coverslip glass-bottom , multi-well plates or 35 mm dishes ( MatTek ) . After an additional 24 hr , cells were imaged on a Leica DMi8 inverted microscope , equipped with a 488 excitation laser , Yokogawa CSU-W1 spinning disk confocal scanner , 63x Plan Apo 1 . 4 NA oil-immersion objective , and controlled by Metamorph software . Environmental control was maintained using a Tokai Hit stagetop incubator . DIC and GFP images were collected every 30 seconds with a Hammatsu Orca Flash 4 . 0 CMOS camera . For phase contrast videomicroscopy , stable-expressing mCherry-H2B Hela were cultured and transfected as above prior . Epifluorescence and phase contrast videomicroscopy were captured on the Nikon Eclipse microscope as above with a humidified incubator to maintain cells at 37°C with 5% CO2 . Images were captured every 5 minutes with phase contrast and mCherry epifluorescence and mitotic cells were analyzed by visualizing furrows . Quantitative immunofluorescence was performed to evaluate GFP-Plk1 levels at equatorial membrane versus poles with regions of interest ( ROIs ) as shown in Figs 3C and 4H . ROIs were chosen internal to the external membrane to minimize the confounding effects of intra-versus extra-cellular levels of GFP-anillin . Membrane lipid arrays ( P-6002 , Echelon Biosciences ) were incubated with 1 μg/ml GST fusion proteins in PBS containing 0 . 1% Tween-20 and 3% BSA for 3 h at 25°C . After washing , proteins were detected using anti-GST antibody . Membranes were imaged with the Odyssey infrared imaging system ( LI-COR Biosciences ) and quantified using Image Studio Lite v5 . 2 software ( LI-COR Biosciences ) . To quantify each band a box was drawn around the band to calculate the total pixel intensity within that box . To account for background fluorescence , a second box of equal size was drawn within the same lane and the pixel intensity of that background box was subtracted from the pixel intensity of the box containing the band of interest . Anillin intensities were normalized to β-actin for each sample , then compared to the control knockdown/addback condition to determine the percent anillin knockdown and relative total anillin expression following each transfection . The mean and standard deviation of three independent replicates are reported . Replicate experiments were performed and analyzed by 2-tailed t-test with the comparisons as shown . There were no corrections for multiple t-tests . For multiple comparisons ANOVA was used as described in legends . | Human diseases such as cancer and congenital trisomies arise from loss of genetic material during cell division . Yet in most divisions , cells preserve their genetic integrity by strict coordination of cell membrane cleavage ( cytokinesis ) , with accurate separation of genetic material ( mitosis ) . Thus , understanding how mitosis and cytokinesis are coordinated can provide insight into human disease . Anillin is one central integrator of cytokinesis that is recruited in a ring-like structure on the membrane of dividing cells . Protein phosphorylation is a common mechanism regulating timing of events in mitosis; although 46 phosphorylation sites have been mapped on anillin , their functional significance is unknown . Here , we evaluated the effect of blocking anillin phosphorylation on human cell division . Surprisingly , most phosphorylation events appeared dispensable for cytokinesis in the assay used . By contrast , phosphorylation of S635 is important for early recruitment of anillin to the midzone membrane , furrow stabilization and efficient cytokinesis . Our findings highlight a central mechanism regulating the timing of human cytokinesis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"phosphorylation",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"anaphase",
"hela",
"cells",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"immunology",
"biological",
"cultures",
"mitosis",
"immunoprecipitation",
"cytokinesis",
"cell",
"cultures",
"antibodies",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"immune",
"system",
"proteins",
"chromosome",
"biology",
"proteins",
"cell",
"lines",
"cell",
"membranes",
"precipitation",
"techniques",
"biochemistry",
"cell",
"biology",
"post-translational",
"modification",
"physiology",
"biology",
"and",
"life",
"sciences",
"cultured",
"tumor",
"cells"
]
| 2017 | Anillin Phosphorylation Controls Timely Membrane Association and Successful Cytokinesis |
The identification of the H3K4 trimethylase , PRDM9 , as the gene responsible for recombination hotspot localization has provided considerable insight into the mechanisms by which recombination is initiated in mammals . However , uniquely amongst mammals , canids appear to lack a functional version of PRDM9 and may therefore provide a model for understanding recombination that occurs in the absence of PRDM9 , and thus how PRDM9 functions to shape the recombination landscape . We have constructed a fine-scale genetic map from patterns of linkage disequilibrium assessed using high-throughput sequence data from 51 free-ranging dogs , Canis lupus familiaris . While broad-scale properties of recombination appear similar to other mammalian species , our fine-scale estimates indicate that canine highly elevated recombination rates are observed in the vicinity of CpG rich regions including gene promoter regions , but show little association with H3K4 trimethylation marks identified in spermatocytes . By comparison to genomic data from the Andean fox , Lycalopex culpaeus , we show that biased gene conversion is a plausible mechanism by which the high CpG content of the dog genome could have occurred .
Until recently the mechanisms controlling the localization of recombination hotspots in mammalian genomes were largely unknown . However , recent research has revealed that zinc-finger protein , PRDM9 , binds to specific DNA motifs in the early stages of recombination initiation in order to direct such events [1]–[3] . PRDM9 trimethylates lysine 4 of histone H3 ( H3K4me3 ) , an epigenetic modification specifically enriched around recombination initiation sites [4]–[6] . The importance of PRDM9 for the localization of recombination events has been demonstrated in both humans [7] , [8] and mice [2] , [6] , with recent results in mice suggesting that PRDM9 determines the location of virtually all recombination hotspots in these organisms [6] . Variation in the zinc-finger encoding domain of PRDM9 in humans can alter the DNA motif to which the protein binds , and in turn alter the activity of recombination hotspots [7]–[9] . High levels of variation in PRDM9 across species [2] , [10] , [11] may explain why humans and chimpanzees do not share recombination hotspots despite very high levels of overall sequence identity [10] . Indeed , PRDM9 shows clear evidence of rapid evolution across the metazoan taxa covering an era of roughly 700 million years [11] . Despite successful formation of double-strand breaks ( DSBs ) [6] , the initiating event in meiotic recombination , knockout of Prdm9 results in infertility in both male and female mice with arrest of spermatogenesis and oogenesis at pachynema , impairment of DSB repair , chromosome asynapsis , and disrupted sex-body formation in males [12] . Intriguingly , Prdm9 has been shown to be involved with hybrid sterility , potentially implicating it in the process of speciation [12] . Given the wide-ranging importance of PRDM9 , it was therefore surprising to note that dogs ( Canis familiaris ) and other canids are the only known mammals to carry functionally inert versions of PRDM9 with multiple disruptive mutations [11] , [13] . This implies that either the function of PRDM9 is carried out by another gene , or that dogs have been able to avoid the loss of fertility associated with loss of PRDM9 while also ensuring that recombination continues to occur . In order to gain insight into recombination across the canine genome , we have constructed a genetic map using patterns of linkage disequilibrium ( LD ) estimated from next-generation sequencing of 51 dogs . Methods for estimating recombination rates from polymorphism data have been validated at both broad and fine scales [14]–[16] , and have previously been used to obtain relatively broad-scale recombination rate estimates in dogs via the use of SNP microarray data [13] . A potential concern when using such methods in dogs is that breed-formation bottlenecks can lead to considerable levels of inbreeding . For this reason , we have utilized genetic polymorphism data collected primarily from free-ranging dogs from geographically diverse regions ( Table S1 ) and largely lacking a history of excessive selective breeding following the original domestication event [17] . These non-breed dogs , which we term ‘village dogs’ , show dramatically reduced levels of homozygosity , and a faster decay of LD when compared to inbred breed dogs ( Figure S1 ) .
In humans and mice , specific DNA motifs have been implicated as the binding sites for PRDM9 [1] , [4] , [21] . In order to investigate if DNA motifs could be identified within canine recombination hotspots , we selected 6 , 228 hotspot regions with no missing sequence data . For each hotspot we identified a region on the same chromosome showing no evidence for local recombination rate elevation ( ‘coldspots’ ) , and with GC content within 0 . 5% of that of the hotspot , and CpG content within 0 . 1% . If more than one such region could be found , we selected the one that matched the hotspot most closely in terms of SNP density . In this way , we were able to identify 4 , 759 hotspots with matched coldspots . Using the sequences of the matched hotspots and coldspots , we performed a search of motifs showing enrichment in hotspot sequences . Our results indicate an extremely strong association with CpG-rich motifs ( Table 1 ) , with the most significant motif being the 7-mer CGCCGCG ( p = 1 . 1e-21 , Fisher's Exact Test after Bonferroni correction ) , which is found in 21 . 3% of hotspots but only 13 . 2% of coldspots , a relative enrichment of 61% . These highly CpG-rich motifs retain significantly high levels of enrichment in hotspots having masked either repeat or non-repeat DNA sequence . Both GC and CpG content show a strong association with canine recombination at fine scales ( Figure S8A and B ) . However , CpG content shows a stronger correlation with recombination rate than GC content over multiple scales ( r = 0 . 37 vs r = 0 . 25 respectively at 1 Mb; Figure S8C and D ) . If both measures are included as predictors in a multiple regression model , CpG content has a positive association , whereas GC content is negative ( Table S3 ) . The influence of GC and CpG content can also be seen when considering the average recombination rate around DNA repeats . The most recombinogenic repeats are low-complexity with high levels of GC and CpG content ( Figure 2A ) . In contrast , the majority of LINE and SINE elements exhibit recombination rates close to the genome average , with a few such as Looper and L1_Canid2 showing weak suppression of recombination . The association between recombination and CpG-dense regions is suggestive of an association with gene promoter regions . Indeed , we observe highly elevated rates of recombination around transcription start sites ( TSS; Figure 2B and Figure S9 ) , dwarfing the elevation that has been observed around TSS in humans and chimps [10] , [22] . Of the 7 , 677 called hotspots , 29% overlap with a TSS , and 50% are within 14 . 7 kb . Only a small fraction ( 14% ) of hotspots appear to be over 100 kb from a TSS . However , the elevation in recombination rate around TSS appears to be associated with CpG islands serving as promoter regions rather than the TSS themselves , as the recombination peak is shifted away from the TSS for genes with the nearest CpG island at some distance from the TSS ( Figure 2C ) , with genes without a nearby CpG island not showing large peaks in local recombination . Conversely , CpG islands containing TSS show elevated recombination rates relative to CpG islands at some distance from TSS ( Figure S10 ) , although interestingly CpG islands >10 kb from the nearest TSS show higher rates than those near ( but not containing ) a TSS . PRDM9 is thought to have been disrupted early in canid evolution , as previous work has shown that the amino acid coding sequence contains multiple disruptive mutations across a diverse set of canid species [11] , [13] . We have further investigated the extent of PRDM9 disrupting mutations within the Canidae family by sequencing within exon containing the zinc-finger domain of PRDM9 in the Lycalopex and Urocyon genera . Within in the Andean fox , Lycalopex culpaeus ( 6–7 . 4 Mya divergence from dogs [23] ) , we found the same disrupting frameshift mutation as has observed in dog ( Table S4 and Table S5 ) , as well as an additional frameshift , and a premature stop codon . In the Island fox , Urocyon littoralis , ( >10 Mya divergence from dogs [23] ) , while we do not observe the same mutations seen in dog , we do observe a distinct premature stop codon , indicating that PRDM9 has been disrupted in this species as well . As none of the identified mutations are common to all species , it would appear that the original disruptive mutation likely occurred outside of sequenced exon . Due to the early loss of PRDM9 , it has been suggested that fine-scale patterns of recombination may be shared across species in the canid lineage [13] , in contrast to other mammalian species in which hotspots are not shared [10] . Such inferences have been based on the effect of Biased Gene Conversion ( BGC ) , in which a recombination-associated heteroduplex in the vicinity of an existing polymorphism can produce base-pair mismatches that are preferentially repaired with C/G alleles rather than A/T alleles ( Figure S13 ) . As such , BGC increases the probability of a C/G allele being transmitted to the next generation , and sustained BGC can ultimately alter the base composition of the genome [24] . To investigate if BGC is active around canine recombination hotspots , we consider the ratio of AT→GC polymorphisms to GC→AT polymorphisms within local regions of the genome ( see Supplementary Text S1 ) . In order to polarize polymorphisms in dog , we sequenced a female Andean Fox , which , as described above , is diverged from dogs by approximately 6–7 . 4 million years [23] and also lacks a functional version of PRMD9 . The fox sample was sequenced to approximately 11× coverage using 100 bp paired-end Illumina sequencing . In the absence of a reference genome for this species , reads were mapped to the dog genome ( canFam3 . 0 ) . Using this data , we polarized the ancestral and derived alleles for polymorphisms observed in dog by assuming that shared alleles represent the ancestral allele . Using this data , we were able to polarize 3 . 2 million polymorphisms on the dog lineage with high confidence . Likewise , we performed SNP discovery in the fox data , and were able to identify and polarize 1 . 2 million polymorphisms on the fox linage . Around canine recombination hotspots , we observe a localized skew in the rate of AT to GC acquisition in the dog genome ( Figure 3A ) , which is stronger for hotspots with higher peak recombination rates ( Figure 3B; Figure S11A ) . The effect remains visible even after excluding all polymorphisms within putative CpG sites ( Figure S11B ) . The dog genome is notable for its high density of CpG-rich regions [25] , [26] . Given CpG-rich regions are highly recombinogenic in the dog genome , it is plausible that BGC is the mechanism by which the dog genome has acquired its high density of CpG rich regions . Specifically , if CpG regions are promoting recombination , and are thereby acquiring increasing levels of GC content via BGC , this would in turn further increase the CpG content of the region and hence further increase recombination in a self-reinforcing process . If dogs and foxes shared the same hotspots , a similar pattern would be expected for polymorphisms observed on the fox linage . However , we do not see evidence of a skew in fox linage polymorphisms around canine recombination hotspots ( Figure 3A ) , which would imply that these two species do not share recombination hotspots . The role of PRDM9 is to trimethylate histone H3K4 , and studies in mice have shown that nearly 95% of hotspots overlap an H3K4me3 mark [4] . It is therefore interesting to ask if canine recombination maintains an association with H3K4me3 , especially given the apparent elevation of recombination around gene promoter regions in dogs . We have used ChIPseq to identify regions of H3K4me3 in dog spermatocytes during the leptotene/zygotene ( L/Z ) and pachytene phases of prophase I of meiosis . We identified 28 , 349 autosomal ChIPseq peaks in L/Z and 32 , 830 for pachynema . Of these , 8 , 721 ( 31% ) peaks were unique to L/Z and 13 , 613 ( 41% ) unique to pachynema . While recombination rates do appear highly elevated around H3K4me3 marks , the effect is almost entirely explained by the presence or absence of a putative CpG island overlapping the mark ( Figure 4A and B ) . The pattern is very similar for both L/Z and pachytene cells , albeit with a small increase in rate for pachytene-specific marks in the absence of CpG islands ( Figure S12 ) . Conversely , while CpG islands overlapping H3K4me3 marks are ∼60% more recombinogenic than those without marks , a strong elevation in local recombination rate remains visible for islands without H3K4me3 marks . Notably , putative CpG islands without H3K4me3 marks also appear to have elevated background rates ( Figure 4C ) , and this effect persists even after extensive thinning CpG islands to ensure lack of clustering . This could reflect background sequence context , as islands with H3K4me3 marks have lower levels of CpG content in the flanking 50 kb than those without marks ( 2 . 2% vs 1 . 2% , p≪1e-16 ) , or may be indicative of other epigenetic factors such as DNA methylation .
The apparent loss of PRDM9 in canids makes dogs of particular interest for the study of meiotic recombination . Our study reveals that while broad scale patterns of recombination appear superficially typical of mammals , dogs appear to have a quite different landscape compared to other mammals so far studied at the fine scale . Of particular note is the strong association of recombination with CpG-rich features of the genome , particularly around promoter regions , which is reminiscent of the double-strand break localization around H3K4me3 marks that has been observed around promoter regions in Prdm9-knockout mice [6] . However , in contrast to these results , we find that the elevation in canine recombination rate around promoters appears to be primarily associated with CpG content , and shows little association with H3K4me3 marks . The association with promoter regions is also superficially reminiscent of the elevated recombination rates observed around promoter in yeast , which has been related to nucleosome spacing [27] , [28] . The biased conversion of A/T to G/C alleles at already CpG enriched recombination hotspots may help explain the 2–3-fold increase in putative CpG islands in the dog genome compared to human and mouse [25] . Despite the relative high density of CpG islands in the dog genome , it has also been noted that dogs have fewer promoter-associated CpG islands than humans or mice , especially near essential and highly expressed genes [25] . If the role of PRDM9 is to deflect DSBs away from functional elements , as has been suggested in other species [6] , then the preferential loss of recombinogenic CpG islands near promoters in dogs may indicate that selection is acting to deflect recombination from genic regions since the loss of PRDM9 . PRDM9-knockout mice are infertile due to the failure to properly repair DSBs [29] . How canids escaped this infertility is unknown , but it must have occurred in the common ancestor of dogs , wolves , foxes and jackals , but after the split from Panda ∼49 Mya [13] . The stable recombination landscape of canids resulting from the loss of PRMD9 may contribute to the ability to successfully hybridize relatively divergent canine species ( e . g . dog and jackal [30] ) . Our comparison of substitution patterns in dogs and Andean foxes does not support the hypothesis that the death of PRDM9 has resulted in the evolutionary stability of recombination along the canid lineage , at least at the fine scale . It is therefore possible that while PRDM9 is dysfunctional , an unknown ortholog of PRDM9 could have assumed a similar role . Nonetheless , the observed recombination landscape in dogs does appear to have some unusual features , and it is plausible that these result directly from the loss of PRDM9 . However , it is worth noting that in addition to the loss of PRDM9 , canines are considerably diverged from other species that have so far been studied for fine-scale recombination rate variation , and it is plausible that the observed differences in the recombination landscape have also been influenced by other factors such as genomic structure . In order to fully understand the dynamics of recombination rate evolution , it will be necessary to obtain high-quality and fine-scale genetic maps across a wide range of species . As such maps become available , it will be possible to place the canine map into a proper evolutionary context , and thereby identify the factors that determine the forces that shape the distribution of recombination in the genome .
Canine blood samples were collected under Cornell Institutional Animal Care and Use Committee ( IACUC ) approval ( #2005-0151 , 2007-0076 and 2011-0061 ) in accordance with applicable federal , state , and local laws , regulations , policies , and guidelines . For each animal , approximately 3–5 ml of blood was drawn from the cephalic vein into K2-EDTA blood collection tubes . Dogs were unsedated to minimize handling time and overall distress during the short blood drawing procedure . Culpeo blood was extracted from a captive female fox using a similar procedure during a routine physical by trained veterinary personnel at the Universidad de San Antonio Abad Zoo in Cusco . | Recombination in mammalian genomes tends to occur within highly localized regions known as recombination hotspots . These hotspots appear to be a ubiquitous feature of mammalian genomes , but tend to not be shared between closely related species despite high levels of DNA sequence similarity . This disparity has been largely explained by the discovery of PRDM9 as the gene responsible for localizing recombination hotspots via recognition and binding to specific DNA motifs . Variation within PRDM9 can lead to changes to the recognized motif , and hence changes to the location of recombination hotspots thought the genome . Multiple studies have shown that PRDM9 is under strong selective pressure , apparently leading to a rapid turnover of hotspot locations between species . However , uniquely amongst mammals , PRDM9 appears to be dysfunctional in dogs and other canids . In this paper , we investigate how the loss of PRDM9 has affected the fine-scale recombination landscape in dogs and contrast this with patterns seen in other species . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| []
| 2013 | Genetic Recombination Is Targeted towards Gene Promoter Regions in Dogs |
Leishmania parasites are transmitted in the presence of sand fly saliva . Together with the parasite , the sand fly injects biologically active salivary components that favorably change the environment at the feeding site . Exposure to bites or to salivary proteins results in immunity specific to these components . Mice immunized with Phlebotomus papatasi salivary gland homogenate ( SGH ) or pre-exposed to uninfected bites were protected against Leishmania major infection delivered by needle inoculation with SGH or by infected sand fly bites . Immunization with individual salivary proteins of two sand fly species protected mice from L . major infection . Here , we analyze the immune response to distinct salivary proteins from P . papatasi that produced contrasting outcomes of L . major infection . DNA immunization with distinct DTH-inducing salivary proteins from P . papatasi modulates L . major infection . PpSP15-immunized mice ( PpSP15-mice ) show lasting protection while PpSP44-immunized mice ( PpSP44-mice ) aggravate the infection , suggesting that immunization with these distinct molecules alters the course of anti-Leishmania immunity . Two weeks post-infection , 31 . 5% of CD4+ T cells produced IFN-γ in PpSP15-mice compared to 7 . 1% in PpSP44-mice . Moreover , IL-4-producing cells were 3-fold higher in PpSP44-mice . At an earlier time point of two hours after challenge with SGH and L . major , the expression profile of PpSP15-mice showed over 3-fold higher IFN-γ and IL-12-Rβ2 and 20-fold lower IL-4 expression relative to PpSP44-mice , suggesting that salivary proteins differentially prime anti-Leishmania immunity . This immune response is inducible by sand fly bites where PpSP15-mice showed a 3-fold higher IFN-γ and a 5-fold lower IL-4 expression compared with PpSP44-mice . Immunization with two salivary proteins from P . papatasi , PpSP15 and PpSP44 , produced distinct immune profiles that correlated with resistance or susceptibility to Leishmania infection . The demonstration for the first time that immunity to a defined salivary protein ( PpSP44 ) results in disease enhancement stresses the importance of the proper selection of vector-based vaccine candidates .
In leishmaniasis , phlebotomine sand flies transmit Leishmania parasites to a mammalian host by depositing the parasite in the skin during probing and feeding . Together with the parasite , sand flies deposit a repertoire of salivary components that assist the sand fly in getting a blood meal [1] . Some of these salivary proteins are immunogenic in humans , canids and mice [2]–[5] . Repeated exposure to sand fly salivary gland homogenate ( SGH ) or sand fly bites have been shown to protect mice to subsequent challenge with Leishmania major and SGH [6] or L . major infected sand flies [7] . The protective effect of insect saliva is not exclusive to sand flies and leishmaniasis . Animals pre-exposed to tick bites were protected from Borrelia infection [8] and from the fatal outcome of tularemia [9] . Moreover , immunization with a single tick salivary protein protected mice from the fatal outcome of encephalitis virus [10] . Furthermore , pre-exposure to mosquito bites protected mice against Plasmodium berghei infection [11] and more recently , immunization with the saliva of an aquatic insect ( Naucoris genus ) protected animals against Mycobacterium ulcerans infection [12] . To date , only two sand fly salivary proteins , Maxadilan from Lutzomyia longipalpis and PpSP15 from Phlebotomus papatasi , have shown promise as protective molecules against leishmaniasis [13] , [14] . It is proposed that immunity to maxadilan neutralizes exacerbation of L . major infection [13] , while immunization with PpSP15 results in protection of wild-type and B-cell deficient mice indicating that cellular immunity to PpSP15 is sufficient for protection [14] . Moreover , the protection observed by immunization with PpSP15 was associated with a DTH response [14] . More recently , Oliveira et al . investigated the IgG isotypes produced by DNA immunization with plasmids encoding distinct DTH-inducing sand fly salivary proteins and showed that some molecules produce IgG2a antibodies indicative of a Th1 response while others surprisingly produced IgG1 , a marker for Th2 response in mice [15] . In this work we identified two additional DTH-inducing salivary proteins in P . papatasi , PpSP42 and PpSP44 . Mice immunized with either of these molecules were not protected against L . major infection . Moreover , PpSP44-immunized mice showed aggravated lesions . This allowed us to explore how immunity to specific salivary proteins could affect the outcome of L . major infection . We show for the first time that an early adaptive immune response specific to a salivary protein is able to prime the anti-Leishmania immune response leading to protection or exacerbation of L . major infection . More importantly , this adaptive response is efficiently elicited by sand fly bites , the natural route of transmission .
P . papatasi Israeli strain sand flies were reared at the Walter Reed Army Medical Research Institute and at the Laboratory of Malaria and Vector Research , NIAID , NIH , as described elsewhere [14] . Preparation of salivary gland homogenate ( SGH ) and pre-exposure of mice ( Charles River Laboratories Inc ) to uninfected sand flies was carried out according to Valenzuela et al . [14] and Kamhawi et al . [7] . Experiments were performed using 6 to 8 weeks old C57BL/6 mice under pathogen free conditions . All animal studies were approved by the Animal Care and Use Committee at The National Institute of Allergy and Infectious Diseases . Ten DNA plasmids encoding to P . papatasi salivary gland-secreted proteins were cloned into the VR2001-TOPO vector and purified as previously described [15] . Mice were immunized intradermally in the right ear three times at two weeks intervals with 5 µg of DNA plasmid in 10 µl sterile water or with the equivalent of 0 . 5 sand fly salivary gland pairs in 10 µl PBS [15] . Two weeks after the last DNA immunization , animals were challenged intradermally in the left ear with P . papatasi SGH ( 0 . 5 salivary gland pair/10 µl ) to test for DTH inducing salivary proteins . For infection , a mixture of 0 . 5 pairs SGH and 500 L . major metacyclics in 10 µl ( SGH-LM ) was used to mimic the natural route of transmission . L . major clone V1 ( MHOM/IL/80/Friedlin ) was cultured in 199 medium with 10% heat-inactivated fetal bovine serum ( HyClone ) , 100 U/ml penicillin , 100 µg/ml streptomycin , 2 mM L-glutamine and 40 mM Hepes . The ear thickness was measured 48 hours following intradermal injection of P . papatasi SGH . Values are represented as Δ ear thickness ( ear thickness of experimental groups subtracted from the mean ear thickness of naïve mice 48 hours after injection with 0 . 5 pair of SGH ) . For measurements of Leishmania lesions , the largest diameter was recorded on a weekly basis . Ear thickness and lesion diameter were measured using a Digimatic caliper ( Mitutoyo Corp . ) . Total genomic DNA was extracted from mice ears using the DNeasy tissue kit following the manufacturer's protocol ( Qiagen ) . A total of 100 ng was amplified by real time PCR ( LightCycler 480 , Roche Diagnostics ) using primers JW11 and JW12 [16] and 18S primers as a housekeeping gene with the FastStart Sybr green I kit ( Roche ) . The standard curve was generated using DNA from naïve ears spiked with 10-fold serial dilutions of L . major DNA . Expression levels were normalized to 18S DNA and corrected for the weight of the whole ear . Values represent the relative number of parasites per ear . Cells were recovered from the ear dermis as described previously [6] . Cells ( 5×106 ) were stimulated with or without 100 µg soluble Leishmania antigen ( SLA ) for 12 hours . The cells were then stimulated with 20 ng PMA and 500 ng ionomycin , in the presence of monensin ( 2 µM final concentration ) for 4 hours . For surface markers , cells were washed , incubated for 15 min at 4°C with 2 . 4G2 mAb to block FcγR , and stained with APC-Cy7 αCD4 ( RM4-5 ) and APC-TCRβ chain ( H57-597 ) for 20 min at 4°C . The cells were fixed , permeabilized ( Cytofix/Cytoperm Plus; BD Pharmingen ) and stained with PE-Cy7 αIFN-γ ( XMG 1 . 2 ) and PE αIL-4 ( 11B11 ) . The data were collected using a FACSArray ( BD Biosciences ) and analyzed with FlowJo software ( Tree Star ) . The lymphocytes were gated using size , granularity and surface markers . Expression profile of cytokines , chemokines , and related inflammatory genes was generated using the mouse inflammatory cytokines and receptor Oligo GEArray ( OMM-011; Superarray ) . This array contains 112 genes representing cytokines , receptors and housekeeping genes . Two hours after challenge , total RNA was isolated from the left ears using QIAshredder ( Qiagen ) and RNeasy Mini Kit ( Qiagen ) following the manufacturer's instructions . RNA ( 6 µg ) from a pool of seven ears was amplified and labeled with biotin 16-UTP ( Roche Diagnostics ) using the SuperArray TrueLabeling-RT Enzyme kit ( Superarray ) . The resulting biotinylated cRNA was hybridized overnight to the Oligo GEArray® membrane . After washing and blocking the array membranes , alkaline phosphatase-conjugated streptavidin was added to the membrane followed by CDP-Star substrate . A chemiluminescent signal was acquired using the Image Station 2000 MM ( Kodak ) . The data was analyzed using the GEArray Expression Analysis Suite ( Superarray ) . Analysis parameters were set to local background correction and normalized to a set of housekeeping genes included in each membrane . Results were expressed as the fold increase in the intensity of the captured signal over the levels in naïve ears challenged with SGH-LM . Only genes showing a four-fold or higher change in expression compared to the naïve group in at least two of three independent experiments were considered . The genes that showed a four-fold or higher change in expression over control using the GEArray were validated by Real time PCR . Five µg of total RNA from mice ears was used for the synthesis of cDNA ( Superscript III , Invitrogen ) following the manufacturer's instructions . The cDNA was amplified with the 480 Master SYBR Green I mix ( Roche Diagnostics ) and gene specific primer sets for IFN-γ , IL-4 , IL-5 , TNF-α and IL-12Rβ2 ( Superarray ) using the LightCycler 480 ( Roche Diagnostics ) . A standard curve for each set of primers was generated as recommended by the manufacturer . The expression levels of the genes of interest were normalized to endogenous 18S RNA levels . The results are expressed in fold change over naive ears challenged with SGH-LM . Statistical evaluation of the means of experimental groups was done using one-way analysis of variance followed by the Tukey-Kramer post-test . Data from parasite numbers were log transformed before conducting statistical tests . Significance was determined as p<0 . 05 . All statistical tests and graphs were done using Prism-GraphPad version 5 ( GraphPad Software Inc . ) .
Of 10 different DNA plasmids coding for the most abundant P . papatasi salivary proteins [14] , [17] , PpSP12 ( 12-kDa protein; AF335485 ) , PpSP14 ( 14-kDa protein; AF335486 ) , PpSP15 ( 15-kDa protein; AF335487 ) , PpSP28 ( 28-kDa protein; AF335488 ) , PpAg5 ( 29-kDa protein; ABA54266 ) , PpSP30 ( 30-kDa protein; AF335489 ) , PpSP32 ( 32-kDa protein; AF335490 ) , PpSP36 ( 36-kDa protein; AF261768 ) , PpSP42 ( 42-kDa protein; AF335491 ) , and PpSP44 ( 44-kDa protein; AF335492 ) , only mice immunized with PpSP15 , PpSP42 and PpSP44 DNA plasmids showed a statistically significant ( p<0 . 05 ) DTH response 48 hours following challenge with SGH as measured by Δ ear thickness compared to control DNA-immunized mice ( CTL DNA ) ( Fig . 1 ) . However , immunization with PpSP12 , PpSP14 , PpAg5 , PpSp32 and PpSP36 produced humoral responses ( data not shown ) indicating in vivo expression of the corresponding proteins . PpSP15 is a 15 kDa salivary protein of unknown function present only in sand flies [14] , [17] . PpSP42 and PpSP44 are salivary proteins that belong to the Yellow family of proteins [17] with a predicted molecular weight of 42 and 44 kDa respectively . Immunization with PpSP15 DNA or protein was previously shown to produce a DTH response and to protect animals from L . major infection [14] . Here we reaffirm the protective nature of PpSP15 but show that immunization with PpSP42 and PpSP44 , the remaining DTH-inducing molecules , do not confer protection against L . major infection ( Fig . 2 ) . As predicted SGH or pre-exposure to uninfected sand fly bites also control L . major infection up to nine weeks post-challenge ( Fig . 2 ) . Mice immunized with PpSP44 exacerbated the infection showing progressive lesions that were predominantly ulcerative . The lesion size in this group was not measured beyond week seven due to extensive tissue damage ( Fig . 2 ) . This group was chosen for comparison to protected PpSP15-immunized mice for a better understanding of the contribution of anti-saliva immunity to the course of Leishmania infection . The parasite load was investigated at 2 , 6 , 9 and 11 weeks post-infection in PpSP15- and PpSP44-immunized mice . By 6 weeks post-infection , a significant decrease in parasite load was observed in mice immunized with PpSP15 compared with control DNA or PpSP44-immunized mice ( data not shown ) . PpSP15-immunized mice maintained a 3 log reduction in parasite load up to 11 weeks post-infection . Panels A–C show representative ears of PpSP44- , PpSP15- and control DNA-immunized mice , respectively , 11 weeks post-infection ( Fig . 3 ) . Overall , the ears of PpSP15-immunized mice ( Panel B ) showed little to no tissue damage while those of PpSP44-immunized mice showed severe tissue erosion ( Panel A ) . The ears of mice immunized with control DNA ( Panel C ) were intermediate showing ulcerated lesions with moderate tissue damage . Interestingly , the parasite loads were comparable in mice immunized with PpSP44 and control DNA , suggesting that the number of parasites in the ear of PpSp44-immunized animals was not entirely responsible for the extensive damage observed in these animals . The observed protection and exacerbation of L . major infection in PpSP15- and PpSP44-immunized mice , respectively , correlates with the expression of IFN-γ and IL-4 by CD4+ T cells recovered from the ears of these mice two weeks after challenge with SGH-LM ( Fig . 4 ) . Following in vitro stimulation with soluble Leishmania antigen ( SLA ) , 31 . 5% of CD4+ T cells in PpSP15-immunized mice produced IFN-γ compared to only 7 . 1% and 7 . 8% in mice immunized with PpSP44 and control DNA respectively ( Fig . 4 , top panels ) . IL-4 production was low in PpSP15-immunized mice ( 2 . 5% of CD4+ T cells ) . In comparison , 8 . 2% and 6 . 3% of CD4+ T cells produced IL-4 in mice immunized with PpSP44 and control DNA , respectively ( Fig . 4 , bottom panels ) . These data suggest that the immune response to distinct salivary proteins has a polarizing effect on the outcome of Leishmania infection . To understand the basis of the different outcomes of L . major infection in mice immunized with PpSP15 and PpSP44 we compared the early mRNA expression profiles of the inflammatory cytokines in the ears of these mice two hours following challenge with SGH-LM . Using the “Inflammatory Cytokines and Receptors” macroarray , transcripts showing a four-fold or higher change in signal intensity of gene expression compared to naïve controls were further analyzed and are presented in Table 1 . PpSp15-immunized mice consistently produced high levels of IFN-γ and IL-12-Rβ2 and low levels of IL-4 and IL-5 ( Table 1 ) . In contrast , PpSP44-immunized mice produced high levels of IL-4 and IL-5 and baseline levels of IFN-γ transcripts . TNF-α transcripts were present at relatively high levels in mice immunized with PpSP15 and PpSP44 ( Table 1 ) . Real-time PCR was used to validate the results of the macroarray and showed that PpSP15-immunized animals induced a three-fold increase in IFN-γ and IL-12-Rβ2 messages compared to mice immunized with PpSP44 ( p<0 . 05 ) ( Fig . 5 ) . Conversely , mice immunized with PpSP44 showed a 20-fold increase in the expression of IL-4 ( p<0 . 005 ) and no significant expression of IFN-γ and IL-12-Rβ2 ( Fig . 5 ) . No significant difference was observed in the expression of IL-5 or TNF-α . The amount of each salivary protein injected by sand flies during feeding is unknown . Therefore , we investigated whether the early induction of IFN-γ and IL-4 in mice immunized with PpSP15 and PpSP44 , observed by challenge with SGH-LM , is reproducible by challenge with sand fly bites . In addition , uninfected sand flies were used to demonstrate that this response remains unchanged in the absence of parasites . Two hours following uninfected sand fly bites , mice immunized with PpSP15 showed a three-fold higher expression of IFN- γ and a five-fold lower expression of IL-4 compared with PpSP44-immunized mice ( Fig . 6 ) . There were no significant differences in the expression of IL-12Rβ2 or IL-5 amongst mice immunized with PpSP15 , PpSP44 and control DNA ( data not shown ) . This response shows that an adaptive immune response specific to distinct salivary proteins is inducible as early as two hours following sand fly bites and that the amount of salivary protein injected by the bite of a sand fly is sufficient to produce a specific and strong recall response in immunized animals .
It is established that a Th1 immune response and the production of IFN-γ are correlated with protection from L . major infection in C57BL/6 mice [18] . Conversely , a Th2 immune response is associated with susceptibility [18] . Earlier studies have demonstrated the potential of immunity to sand fly saliva in the control of Leishmania infection [6] , [7] , [13] , [14] . More information is needed to define the immune profile induced by distinct salivary proteins and its specific effect on the outcome of disease . In this work , we demonstrate that DTH-inducing P . papatasi sand fly salivary molecules are not universally protective against L . major infection and that immunity to some can result in its exacerbation . Mice immunized with PpSP15 controlled the infection and had significantly lower parasite load compared to naïve mice , as previously reported [14] . In contrast , mice immunized with PpSP44 exacerbated the infection showing lesions with severe tissue erosion and maintaining a high number of parasites up to 11 weeks post-infection . This is the first account in which an immune response to a defined sand fly salivary protein results in disease exacerbation . Protection in PpSP15-immunized mice and exacerbation in PpSP44-immunized mice were correlated with an anti-Leishmania Th1 and Th2 immune response , respectively ( Fig . 4 ) . The anti-Leishmania immune response was characterized by a considerable increase in IFN-γ producing CD4+ T cells in PpSP15-immunized mice ( over four-fold higher compared to control DNA- and PpSP44-immunized mice ) and over three-fold lower IL-4 producing CD4+ T cells compared to PpSP44-immunized mice ( Fig . 4 ) . At this time point a small increase in the percent of CD4+ T cells producing IL-4 in PpSP44-immunized mice was detected compared to controls . Nevertheless , there is clear exacerbation both in lesion size and tissue pathology in PpSP44-immunized mice ( Fig . 2 , 3 ) . We propose that the polarization of anti-Leishmania immunity towards a Th1 or Th2 response in these mice is the result of their prior immunization with DNA encoding the respective salivary proteins . Earlier studies have hypothesized that anti-saliva immunity leads to protection from L . major by the creation of a hostile environment that kills the parasite , acceleration and priming of the anti-Leishmania immunity , or a combination of both [7] , [14] . Indeed , mice protected from L . major infection through pre-exposure to sand fly bites showed an increase in the frequency of ear epidermal cells producing IFN-γ and IL-12 six hours after challenge [7] . This rapid production of IFN-γ prompted us to investigate the expression profile of pro-inflammatory cytokines induced by PpSP15 and PpSP44 at an early time point ( two hours ) following challenge with SGH-LM . Macroarray results validated by real-time PCR showed that mice immunized with PpSP15 selectively induced transcripts associated with a Th1 immune response ( IFN-γ and IL-12rβ2 ) and downregulated Th2 associated transcripts ( IL-4 ) . IL-12rβ2 is expressed on both activated Th1 CD4+ cells and NK cells [19]–[21] . Recently , it has been shown that NK cells could play a role in adaptive immunity [22] and may be the source of the early IFN-γ expression seen in PpSP15-immunized mice . Alternately , we cannot exclude the possibility that the up-regulation of IFN-γ expression is by specific CD4 memory T cells that are rapidly recruited to the site of infection . The cells that are responsible for the expression of IFN-γ at this early time point is currently under investigation . PpSP44- immunized mice that exacerbated L . major infection selectively induced IL-4 ( a marker of Th2 differentiation ) and did not upregulate IFN-γ showing the specificity of the observed immune responses to each of the salivary proteins . It should be noted that neither IFN-γ nor IL-4 were induced in the CTL DNA-immunized mice . Enhancement of Leishmania infection in mice pre-exposed to sand fly saliva was recently demonstrated for Lu . intermedia and L . braziliensis in a BALB/c model of infection [23] . Mice immunized with SGH of Lu . intermedia showed a low IFN-γ to IL-4 ratio that correlated with an enhanced disease profile [23] . It is possible that the immunodominant protein in the salivary repertoire of Lu . intermedia induces an immune response similar to that of PpSP44 resulting in the exacerbation of L . braziliensis infection in BALB/c mice . This is in contrast to the protection from L . major infection observed in BALB/c mice pre-exposed to P . papatasi bites or SGH [6] , [7] . Interestingly , the molecular weight of a strongly antigenic salivary protein of Lu . intermedia is 45 kDa [23] corresponding to the molecular weight of PpSP44 . This raises the question whether immuno-dominance of salivary proteins vary in different sand fly species . Saliva is composed of a repertoire of proteins and their overall effect is likely influenced by the sand fly species , the Leishmania species and the mammalian host resulting in an overriding exacerbative or protective immune profile . Stimulatory and suppressive immune responses to salivary molecules have been previously described in ticks [24] . Lymphocytes from tick resistant donors proliferated in response to tick salivary gland antigens demonstrating antigen-specific stimulation . However , their non-specific PHA-induced proliferation was significantly suppressed [24] . Since Leishmania is transmitted by sand fly bites we wanted to verify if the small amount of PpSP15 or PpSP44 injected by sand flies during feeding is able to recall the same level and type of immunity observed in response to challenge with SGH-LM . Moreover , we used uninfected sand flies to investigate whether this response is specific to the salivary molecules and is not influenced by the presence of Leishmania parasites in SGH-LM . Sand fly bites induced an early up-regulation of IFN-γ in PpSP15-immunized mice suggesting that this salivary protein can recall a protective Th1 response by the natural route of exposure . PpSP44-immunized mice also reproduced the response observed following challenge with SGH-LM and maintained a high expression of IL-4 and a low expression of IFN-γ ( Fig . 6 ) . Despite the fact that the above responses were elicited by uninfected sand fly bites , infected flies are expected to inject more saliva as a result of difficulty in feeding and increased probing activity [25]–[27] . This further confirms that an immune response specific to a salivary antigen that generates a DTH response with a Th1 profile is able to confer protection against L . major infection , independent of other confounding factors present in the complex feeding behavior of the sand fly . Recently , Vinhas et al . [28] demonstrated that PBMCs from normal volunteers pre-exposed to the bites of uninfected Lu . longipalpis produced IFN-γ following stimulation with SGH . IFN-γ production was also correlated with killing of L . chagasi parasites in a macrophage-lymphocyte autologous culture [28] . This demonstrates for the first time that humans can mount an anti-saliva cellular immune response that correlates with protection from Leishmania infection and emphasizes the need to identify the molecules in saliva that are responsible for this effect . Currently , there is no evidence that humans will mount a cellular immune response to either PpSP44 or PpSP15 proteins . Further studies are needed to elucidate the role of these salivary proteins in endemic areas . Overall , these data suggest that the early induction of a distinct Th1-type immune response by salivary proteins is important for priming a protective immune response against Leishmania infection . Moreover , a DTH response to saliva or a salivary antigen by itself cannot be considered as a correlate of protection against Leishmania infection . In conclusion , this paper clearly demonstrates that immunization with a particular salivary protein can have a profound modulatory effect on Leishmania infection . We believe that this immunization acts through the differential priming of anti-Leishmania immunity resulting in protection or susceptibility to disease . | In vector-borne diseases , the role of vectors has been overlooked in the search for vaccines . Nonetheless , there is a body of evidence showing the importance of salivary proteins of vectors in pathogen transmission . Leishmaniasis is a neglected vector-borne disease transmitted by sand flies . Pre-exposure to sand fly saliva or immunization with a salivary protein protected mice against cutaneous leishmaniasis . Using DNA immunization we investigated the immune response induced by abundant proteins within the saliva of the sand fly Phlebotomus papatasi . We found that one salivary protein protected while another exacerbated L . major infection , suggesting that the type of immune response induced by specific salivary proteins can prime and direct anti-Leishmania immunity . This stresses the importance of the proper selection of vector-based vaccine candidates . This work validates the powerful protection that can be acquired through vaccination with the appropriate salivary molecule and more importantly , shows that this protective immune response is efficiently recalled by sand fly bites , the natural route of transmission . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
]
| [
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"immunology/immunity",
"to",
"infections"
]
| 2008 | Immunity to Distinct Sand Fly Salivary Proteins Primes the Anti-Leishmania Immune Response towards Protection or Exacerbation of Disease |
Optimists hold positive a priori beliefs about the future . In Bayesian statistical theory , a priori beliefs can be overcome by experience . However , optimistic beliefs can at times appear surprisingly resistant to evidence , suggesting that optimism might also influence how new information is selected and learned . Here , we use a novel Pavlovian conditioning task , embedded in a normative framework , to directly assess how trait optimism , as classically measured using self-report questionnaires , influences choices between visual targets , by learning about their association with reward progresses . We find that trait optimism relates to an a priori belief about the likelihood of rewards , but not losses , in our task . Critically , this positive belief behaves like a probabilistic prior , i . e . its influence reduces with increasing experience . Contrary to findings in the literature related to unrealistic optimism and self-beliefs , it does not appear to influence the iterative learning process directly .
Optimism is known to play an important role in human experience leading to more happiness , greater achievements and better health [1] , although inappropriate optimism can also lead to poor choices [2] . Low optimism is also closely associated with depression and anxiety [3] . Because of this , there has been a recent surge of research interest so as to unveil the underlying mechanisms of optimism , its neural substrate and behavioral consequences [4] . Trait optimism is generally measured using questionnaires , the most common of which is the Life Orientation Test-Revised ( LOT-R ) [5] . The LOT-R is a series of six statements ( and an additional four filler items ) with four positively and four negatively worded items ( e . g . “In uncertain times , I usually expect the best” and “If something can go wrong for me , it will” ) which subjects have to score from 0 to 4 according to how much they agree with it . Optimism is thought to affect cognitive processes in at least two ways . First , it biases one's expectations in a positive direction: while optimists view the glass as being half-full , pessimists might perceive it as half-empty . Formally , such a divergence in the interpretation of the same object could result from the influence of different prior beliefs . Second , optimism also appears to impact learning: optimists sometimes maintain positive beliefs in defiance of what should be strong evidence , such as doctors underestimating the risks of treatments or people continuing to buy lottery tickets . Recent work has shown that this may be due to biases towards more readily learning from “good news” ( i . e . outcomes that are better than expected ) than from “bad news” [6]–[9] . This biased learning could serve as a way to maintain the biases on the beliefs themselves . However , it is not known whether this impact of optimism on learning generalizes to all settings or is only restricted to the domain of personally relevant information and self-beliefs . More generally , how optimism relates to measurable cognitive biases is still poorly understood . To approach these questions , we designed a behavioral task in which positive beliefs about future outcomes as well as learning biases could be quantified in individuals , independently from LOT-R scores and subjective introspection . This paradigm allowed us to disambiguate whether trait optimism functions as a prior belief on the likelihood of future outcomes , as a learning bias , or both .
If optimism really functions like a prior , then its influence should fade the more subjects are given evidence about the association of stimuli and reward . If the amount of evidence is sufficiently large then subjects' performance should become independent of their prior biases . For our Bayesian analysis , this means that the simplest model that would describe their performance is one with a non-informative prior . On the contrary , if optimism affects learning , the difference between optimists and pessimists should be maintained or even amplified with experience . We conducted a control experiment aimed at testing this directly . This experiment also excluded a potential confound in the previous experiment , namely that optimistic subjects might have an a priori preference for fractals . The second experiment was identical to the first one , except for two changes in experimental parameters , designed to reduce the level of uncertainty in the task: A total of 51 new participants ( 28 males and 23 females , age range: 17–46 years old ) participated in this version of the experiment . One subject ( male ) was post-hoc excluded from further analysis , because he did not achieve a 50% performance . In line with our hypothesis , we found that under those conditions , the difference of performance between optimistic and pessimistic subjects disappeared ( Figure 2 ) . No correlation was found between the LOT-R score of individual subjects and the mean of their prior ( r = 0 . 009 , p = 0 . 95; a correlation significantly different from that of experiment 1: Fisher's Z = 2 . 26 , p = 0 . 02; achieved power: 1−β = 0 . 87 assuming the effect size is the same as in experiment 1 ) . The shape of the individual priors extracted from the subjects' performance was always close to a non-informative ( i . e . Jeffrey's ) prior ( α = β = 0 . 5 ) . In fact , in this control experiment , contrary to the main experiment , model comparison ( BIC ) shows that the performance of every single subject was better described by the simpler model in which the prior is chosen to be fixed and non-informative rather than by a prior with flexible α and β ( vs . 45% of the subjects for experiment 1 ) . This suggests that , in this case , subjects were able to correctly take into account the evidence and override their prior expectations: they now behave in a way indistinguishable from that of having unbiased prior beliefs . Furthermore , the reinforcement learning models again failed to account for the data better than the Bayesian models , while supporting similar conclusions: the LOT-R score correlated neither with the learning rates nor with the initial value V0 ( all p>0 . 1 ) . Table 2 and 3 present the group averages of the best-fitting parameters for all the models . In view of these results and so as to test whether the dependency of the bias with level of uncertainty could also be observed in the same group of participants ( vs . between two different groups ) , we also re-analyzed the data of experiment 1 . We compared performances ( % choices ) for the fractals that were “over-observed” ( observed more than 4 times ) compared to the fractals that were “under-observed” ( less than 4 times ) . We tested whether optimists and pessimists differed in their “under-observed” and “over-observed” biases using two sample t-tests . Consistent with our hypothesis , we found that the differences in performances between optimists and pessimists was statistically significant for “under-observed” fractals ( p<0 . 01 ) , but not for the “over-observed” ones ( p = 0 . 135 ) . We used a two-sample , one-tailed t-test to test if one effect is significantly greater than the other , and found that this was the case ( p = 0 . 0017 ) . We finally asked whether optimism could also predict prior beliefs about the likelihood of losses , by repeating the experiment with punishments ( i . e . losses of points ) rather than rewards . The experimental procedure was the same as in Experiment 1 , except for the fact that , here , both the CS and the square stimuli were associated with a probability of punishment ( instead of reward ) , depicted by a cartoon of a sad face . Subjects were now asked to estimate the probability of punishment ci associated with the CS and to avoid punishment when choosing between the CS and the square stimulus . A total of 51 subjects ( 29 males and 22 females , age range: 17–38 years old ) participated in this version of the experiment . Four subjects ( 1 female and 3 males ) were post-hoc excluded from further analysis , because they did not achieve a 50% performance . We found that under those conditions , optimistic and pessimistic subjects had similar performances ( Figure 3 ) . Moreover , subjects' prior mean did not correlate with the LOT-R score ( r = 0 . 049 , p = 0 . 74; significantly different from that of experiment 1 , Fisher's Z = 2 . 05 , p = 0 . 04 and achieved power: 1−β = 0 . 86 ) . The RL models were not as good at explaining the data as the Bayesian model ( in terms of their BIC values , see Methods ) and the extracted model parameters didn't differ between groups ( Table 2 and 3 ) .
In conclusion , trait optimism as measured by the LOT-R questionnaire is found to correlate with performance biases in a simple Pavlovian conditioning task: optimistic subjects over-estimate the probability of reward associated with the uncertain target . This bias affects the estimation of future rewards but not of future losses in our task . It conforms to Bayesian principles of optimal inference and disappears when the level of uncertainty decreases . Our findings are consistent with intuition about the nature of optimism in humans , as well as evidence that optimistic people are more likely than pessimists to have positive gambling expectations [17] . Interestingly in our study the observed estimation biases concern future outcomes for neutral stimuli ( the fractal shapes ) . This can be contrasted with studies looking at unrealistic optimism , which concern self-beliefs . Unrealistic optimism has been defined as the “favorable difference between the risk estimate a person makes for him or herself and the risk estimate suggested by a relevant objective standard” [18] . Compared to such studies , our findings differ in two ways . First , unrealistic optimism studies show that participants are biased in their estimates of positive outcomes ( such as graduating from college , getting married , having a favorable medical outcome ) but even more so in their estimates of negative outcomes ( suffering from a disease , getting divorced etc . ) [18] . In our study , on the other hand , optimism corresponded to an overestimation of the probability of positive outcomes ( reward ) , but not to an underestimation of the probability of negative outcomes ( punishment , experiment 3 ) . This was unexpected at first , since the LOT-R contains statements related to predictions of both positive and negative events . One possible reason might be that the salience of positive and negative outcomes may have differed . Future studies will be needed to assess the generality of the asymmetric effect of optimism , in particular by using more salient negative reinforcers . Second , in our experiment , participants don't seem to be biased in the learning process itself . Optimists and pessimists differ in their initial biases but not in how they accumulate new information . Moreover , fitting the data with reinforcement models showed that they learned similarly from positive prediction errors ( “good news” ) and negative prediction errors ( “bad news” ) . Studies looking at updating of beliefs related to one's personal qualities or future life events [6]–[9] , on the other hand , have typically found that people are likely to discount new information that is worse than their current beliefs , and as such appeared to be “non-Bayesian” learners . For example , Eil and Rao ( 2011 ) find that participants tended to discard negative information ( “bad news” ) when processing personal information regarding their IQ or Beauty , whereas “good news” led to a much tighter adherence to Bayesian updating of their beliefs [7] . Wisfall and Zafar ( 2011 ) also conclude that college students in their study are not Bayesian updaters when they have to form and update their beliefs about their future earnings [9] . Similarly , in a task where participants have to estimate the likelihood of a negative future life event , such as divorce or cancer , Sharot et al ( 2011 ) show that participants updated their beliefs more in response to information that was better than expected compared to information that was worse [6] . There are many important differences between the current paradigm and those studies , which makes the comparison difficult . As stated above , a crucial difference is whether the quantity to be estimated concerns the self or a neutral stimulus . This can lead to large differences in motivation in the learning process: when information is personally relevant , participants have a motive to disregard negative information so that they can keep a rosy view of the future . In our task , on the other hand , there is no intrinsic advantage of keeping a biased estimate for the probability of rewards associated with the fractals . Consistent with this idea , Eil and Rao found that participants conformed Bayesian rationality in their control ( neutral ) condition [7] . Mobius et al . provide a theoretical framework that can possibly unify all these results: they suggest that the updating asymmetry itself can be explained by Bayesian principles in a model where agents derive utility from their beliefs . This model includes the fact that believing that one has a higher than average IQ , for example , even if it is untrue , has an intrinsically “rewarding” value , in that it helps self-confidence [8] . Other differences in experimental design between these studies and ours are worth mentioning . In [6] , for example , the information given about the occurrence probability ( e . g . “actual likelihood cancer 30%” ) is explicit , high-level , provided only once and open to interpretation ( i . e . participants can decide whether this should apply to them or not ) whereas our study involves actual experienced outcomes that have to be integrated over time for the occurrence probability to be estimated . Despite these differences , our results combined with those mentioned above suggest that that there might be at least two distinct computational expressions of optimism: one , corresponding to very general initial biases for simple associations of stimuli and outcomes that can be overcome by learning; and a second one directly affecting the learning process in the domain of personally relevant beliefs with strong emotional content ( such as one's qualities , future health or success ) . It will be important in the future to clarify the boundaries between these domains . The experimental paradigm opens the door to a number of investigations . For example , our experimental paradigm offers new routes to the differentiation between optimism and pessimism , and optimism and hope , which are sometimes believed to be different constructs [19] , [20] . There is a documented link between depression and ( the lack of ) unrealistic optimism [21] , [22] . For example , Strunk et al investigated how participants estimate the likelihood of positive and negative future life events and found that depressed individuals exhibit a pessimistic bias by over-estimating the likelihood of negative future events [22] . It will be important to see how participants with depressive symptoms perform in our task . It will also be interesting to examine the impact of pharmacological manipulations particularly of dopamine or serotonin [23] . Finally , optimistic biases have also been reported in animals and it has been proposed that those biases could be used as an indicator of affective state [24] . For example , Harding et al have found that rats can display optimistic or pessimistic biases when interpreting ambiguous stimuli . Moreover , such biases correlated with the quality of their housing ( unpredictable – which induces symptoms of a mild depression-like state – or predictable ) [25] . In this context , it is interesting that our paradigm can also be adapted for use with animals . It would be very interesting to investigate the relation between cognitive biases observed in such situations of ambiguity with the ones we report . Adapting our paradigm for use with animals will also allows the translational investigation of the underlying neural substrate [26] .
All participants gave informed written consent and the University of Edinburgh Ethics Committee approved the methods used in this study , which was conducted in accordance with the principles expressed in the Declaration of Helsinki . All experiments took place at the Perception lab at the University of Edinburgh . 51 naive subjects took part in each experiment and were recruited mainly among students of the University of Edinburgh . First , subjects were asked to sign a consent form and to fill in the questionnaires . Then , a short trial version of the behavioral task was presented , during which verbal and text instructions were given . Once subjects had confirmed that they were comfortable with the task , they were presented with the full version of the experiment . Visual stimuli were generated using the Matlab programming language and displayed using Psychophysics Toolbox [27]; [28] . Participants viewed the display in a darkened room on a 20″ monitor at a viewing distance of approx . 100 cm . Stimulus sizes on the screen were 8×8 cm and 5×5 cm for fractals and chests respectively . The experiment contained two types of screens ( Figure 1a ) : i ) a series of observation screens which subjects had to passively observe . On each of these screens a fractal stimulus ( or conditioned stimulus , CS ) was shown to be associated with a binary reward ( the presentation of the fractal was followed after 700 ms . by the presentation of a full treasure chest ) or not ( the fractal led to an empty chest ) ; intermixed with ii ) 60 decision screens , where the subject was asked to choose between a fractal stimulus that he had observed before and a blue square , by clicking on it with the mouse . The task of the subjects was to maximize reward gain . More precisely , there were 60 different fractal stimuli in total . They were generated using Matlab code available from the C . I . R . A . M . Research center in Applied Mathematics at the University of Bologna . The probability ci for each fractal CS to lead to a reward was drawn randomly between 0 and 1 at the start of the experiment and kept unknown to the subject . As described above , CSs were then shown in random sequences of observation and decision screens . More precisely , in the main experiment , each CS was assigned to a group of 5 fractals and those were presented in randomly interleaved observation screens before they were shown in decision screens ( Figure 1a – inset ) . In the main experiment , each CS was observed on average 4 times before it appeared on a decision screen ( the exact number was drawn from a Poisson distribution with mean 4 and truncated to be greater than 2 ) . Each CS was involved in only one decision screen . On each decision screen , the reward probability of the square stimulus was drawn randomly from 0 and 1 ( binned with steps of 0 . 1 ) . This probability was explicitly indicated to the subjects , and depicted as a proportion of full circles out of a set of 10 circles ( Figure 1a ) . The side on which the CS appeared in the decision screen was chosen randomly on each trial . Decision screens were displayed until the subject chose one stimulus by clicking on the mouse . The behavioral experiment lasted about 30 minutes . Feedback was not given after each decision screen but each subject was given a final score at the end of the experiment . Due to funding changes , the first 42 subjects of experiment 1 were unpaid but participated in a draw with a £20 voucher prize , while subjects of experiment 2 and 3 and the last 11 subjects in the main experiment were paid £6 for participation ( unrelated to their performance at the task ) . No significant differences were found between paid and unpaid participants' performances . We assumed that subjects behave as Bayesian observers , and estimated the probability of reward , denoted ci , associated with a given fractal i by computing the posterior distribution p ( ci |Di ) , using Bayes rule: ( 1 ) where Di denotes the series of observations related to fractal i ( series of rewards observed , or not , on all observed trials ) and p ( ci ) denotes the subject's prior belief that CS i will be associated with a reward . We further assumed that subjects formed their decision by extracting the mean of this posterior distribution so as to obtain an estimate ĉi of ci: ( 2 ) We modeled the prior distribution p ( ci ) using a beta distribution , which is the conjugate prior of the binomial distribution . This prior has the form: ( 3 ) where Γ denotes the gamma function and parameters α and β control the shape of the prior and are assumed to be the same for all CSs . A prior centered on values lower than chance was considered as a ‘pessimistic prior’ , whereas a prior centered on values greater than chance was considered as an ‘optimistic prior’ in the experiment ( Fig . 1d ) . Under this model , it can be shown that , for each fractal , the posterior mean ĉi is: ( 4 ) where Ni is the number of time fractal i was shown , and ni the number of times it was associated with a reward in the observation screens . We assumed that subjects' decision results from a ‘softmax’ comparison between their estimate ĉi of the probability that the CS should lead to a reward with the probability bi ( explicitly given ) that the square stimulus should lead to a reward on trial t . Subjects would then choose the CS with probability p ( choose fractal ) : ( 5 ) where parameter γ controls how closely the subjects' responses follow the internal estimates and is assumed to be fixed during the whole session . Under this decision-making model , each subject was thus described by 3 free parameters: α , β and γ . These parameters were estimated for each subject based on their task performances , using Maximum Likelihood and numerical optimization methods in Matlab ( fmincon ) . We also fitted various reinforcement learning ( RL ) models to our data . Our idea was to assess whether RL models could capture the differences in performance between optimists and pessimists in experiment 1 , and if so , to identify the parameters which would explain those differences . We were particularly interested in assessing whether optimists and pessimists would differ most in the parameters governing value update as a function of the sign of the prediction error or in those parameters setting the initial biases ( consistent with the alternative account of optimism as a prior belief ) . We used a simple temporal-difference ( TD ) learning algorithm . In these models , subjects learn a value V ( si ) for each CS i , which is initialized at vo ( identical for each CS ) and then updated after each observation of that CS , according to: ( 6 ) where δt = rt−Vt ( si ) denotes the prediction error , rt denotes the binary reward , t represents the observation number , and the learning rate ε ( δt ) is set to hold either the same value ( ε+ = ε− ) for better-than-expected ( i . e . δt>0 ) and worse-than-expected outcomes ( δt<0 ) , or different values ( ε+≠ε− ) . The selection between targets 1 and 2 is governed by a softmax action selection , with additional parameter τ . ( 7 ) where bi corresponds to the reward probability of the colored square . We first examined model RL2b which had 2 free learning rates ε+ , ε− and free vo . We additionally examined simplified versions of this model , which differed in the number of parameters kept free in addition to τ: Each model was fitted to the data of each participant using maximum-likelihood estimation . Table 2 and 3 present the group averages of the best-fitting parameters for all the models . We found that: i ) only the models with free bias term vo captured the difference in performance between optimists and pessimists in experiment 1 ( i . e . led to significantly different parameters for optimists and pessimists ) ; ii ) in line with the hypothesis that optimism functions as a initial bias , the bias vo correlated with LOT-R scores in experiment 1 ( significantly so for RLb;: r = 0 . 541 , p = 0 . 002 ) ; iii ) the RL models were worse at fitting the data than the Bayesian models , both in terms of log likelihood and BIC values in all experiments ( BIC for experiment 1: RLε = 71 . 83; RL2 = 74 . 57; RL2b = 77 . 83; RLb = 75 . 53 , Bayesian model = 60 . 92; BIC for experiment 2: RLε = 80 . 71; RL2 = 83 . 36; RL2b = 86 . 71;RLb = 81 . 16 , Bayesian model = 71 . 38; BIC for experiment 3: RLε = 90 . 52; RL2 = 93 . 88; RL2b = 97 . 49;RLb = 90 . 14 , Bayesian model = 83 . 12 ) . We concluded that , in our data , optimism is well described in terms of a positive prior belief on the likelihood of reward and is not significantly accompanied by selective updating during the learning process . | The optimism bias is regarded as one of the most prevalent and robust cognitive biases documented in psychology and behavioral economics . In individuals , trait optimism is usually measured using self-report questionnaires . However , choices in simple behavioral tasks can also be used to infer how optimistic people are in practice . We asked human subjects to fill in questionnaires about trait optimism , then to participate in a behavioral experiment where they needed to infer the likelihood of visual targets to be associated with a reward . Using modeling , we could then quantify the link between self-report trait optimism and decision or learning biases . We find that people who report that they are optimistic have a positive a priori bias on the likelihood of future reward , whose influence reduces with experience . In our task , trait optimism doesn't distort how new information is integrated: subjects update their estimates similarly following information that is better or worse than expected . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
]
| [
"psychology",
"cognitive",
"psychology",
"social",
"sciences",
"behavior",
"personality",
"biology",
"and",
"life",
"sciences",
"neuropsychology",
"cognitive",
"science",
"neuroscience"
]
| 2014 | Optimism as a Prior Belief about the Probability of Future Reward |
Charcot-Marie-Tooth disease ( CMT ) is a heterogeneous group of peripheral neuropathies with diverse genetic causes . In this study , we identified p . I43N mutation in PMP2 from a family exhibiting autosomal dominant demyelinating CMT neuropathy by whole exome sequencing and characterized the clinical features . The age at onset was the first to second decades and muscle atrophy started in the distal portion of the leg . Predominant fatty replacement in the anterior and lateral compartment was similar to that in CMT1A caused by PMP22 duplication . Sural nerve biopsy showed onion bulbs and degenerating fibers with various myelin abnormalities . The relevance of PMP2 mutation as a genetic cause of dominant CMT1 was assessed using transgenic mouse models . Transgenic mice expressing wild type or mutant ( p . I43N ) PMP2 exhibited abnormal motor function . Electrophysiological data revealed that both mice had reduced motor nerve conduction velocities ( MNCV ) . Electron microscopy revealed that demyelinating fibers and internodal lengths were shortened in both transgenic mice . These data imply that overexpression of wild type as well as mutant PMP2 also causes the CMT1 phenotype , which has been documented in the PMP22 . This report might expand the genetic and clinical features of CMT and a further mechanism study will enhance our understanding of PMP2-associated peripheral neuropathy .
Charcot-Marie-Tooth disease ( CMT ) represents a group of inherited peripheral neuropathies with a heterogeneous genetic and clinical spectrum . The typical CMT phenotype is characterized by distal weakness , sensory loss , foot deformities , and absence of reflexes [1] . CMT is broadly classified into the demyelinating type ( CMT1 ) with slowed motor nerve conduction velocity ( MNCV ) of <38m/s , and the axonal defective type ( CMT2 ) with reduced conduction velocity , >38m/s [2 , 3] . Although the axonal type of CMT is caused by more than 60 genes with diverse mechanisms , the demyelinating type of CMT is caused mainly by peripheral myelin proteins , the salient proteins in the myelin , or their transcription factors [4] . The most frequent genetic cause of CMT1 is alterations in PMP22 , resulting in CMT1A or CMT1E , followed by MPZ mutations , which lead to CMT1B . PMP22 and MPZ account for about 5% and 50% of the total peripheral myelin proteins , respectively . PMP22 mutations account for up to 70–80% of CMT1 cases , while MPZ mutations occur in approximately 10% of CMT1 cases . In addition , a transcription factor for myelin proteins , EGR2 , is the genetic cause of CMT1D [5] . Intriguingly , mutations in Myelin basic protein ( MBP ) and Peripheral myelin protein 2 ( PMP2 ) , which account for 18% and 5% of peripheral myelin proteins , have not been reported as a genetic cause of CMT . MBP is also a constituent of central nervous system ( CNS ) myelin proteins . Although mutation in MBP has not been reported in humans , deletion of the MBP gene causes the ‘shiverer’ phenotype in mice , in which mice show decreased CNS myelination , tremors , and increased severity leading to early death [6] . The PMP2 ( 8q21 . 13 ) was once suspected to be the causative gene because it is located in the close vicinity of the CMT4A locus ( GDAP1 ) [7] , but no mutation was found in the family studied [8] . To link the gene with CMT , a knockout mouse model for PMP2 was generated; however , there was no typical phenotype of peripheral neuropathy except for a slight reduction in the nerve conduction velocity [9] . Recently , a point mutation ( p . I43N ) in PMP2 was strongly suggested as a potential pathogenic mutation in a family with autosomal dominant CMT1 [10] . To demonstrate the pathogenesis of the PMP2 mutation , the researchers showed structural abnormality caused by mutant PMP2 expression in a zebrafish model . Several years ago , we also found a Korean CMT1 family harboring the same PMP2 mutation and have investigated the pathogenicity of the mutation using mouse models . In this report , we present the detailed clinical features of a PMP2 mutation-associated autosomal dominant CMT1 . In addition , the relevance of PMP2 mutation as a genetic cause of dominant CMT1 was assessed using transgenic mouse models .
To determine the genetic cause of the FC183 family , whole exome sequencing was performed on five family members ( S1 Table ) . The mean total sequencing yields was about 8 . 05 Gbp/sample , and the coverage rate of the target region ( ≥10X ) was 91 . 24% . The average number of observed variants per sample was 90 , 653 SNPs and 6 , 299 indels , respectively . Of these , the number of functionally significant variants was 10 , 400/sample . Exome data of the three affected members revealed ~60 functionally significant variants in CMT-related genes ( S2 Table ) . However , none of the variants cosegregated with the affected members in the family . Most variants were polymorphic with high frequency , except for three variants with allele frequencies of less than 0 . 01 in Korean controls . Although three variants ( p . M1I in FAM134B , p . A1315T in SBF2 , and p . A952V in CTDP1 ) were not observed in in Korean controls , they were not considered as the underlying cause , because of noncosegregation with affected individuals in the FC83 family . Since no variants in the CMT-related genes were considered to be an underlying cause of the CMT phenotype , we tried to identify the genetic cause from the entire exome data . A total of 5 cosegregating missense variants were isolated in RYR2 , NME5 , ANK1 , PMP2 , and AGAP11 genes . Given that our family phenotype has a late onset , the age of the proband’s daughter ( IV-1 , 2 years old ) would provide questionable insights into the consideration or exclusion of affection; hence she was excluded from the cosegregation analysis . All five variants were not found in the 500 Korean controls and they were not reported in global human variant databases , such as the dbSNP144 , 1000G , ESP , and ExAC ( Table 1 ) . A thorough look at the variants , their gene functions , conservation , and in silico analysis drew attention to PMP2 ( c . T128A , p . I43N ) mutation . First of all , PMP2 encodes a myelin protein that is an important component of the peripheral nervous system . In addition , the PMP2 mutation cosegregated within extended family members ( Fig 1 ) . The p . I43N mutation in the PMP2 protein is located in the highly conserved lipocalin/cytosolic fatty-acid binding domain , and it was predicted to be pathogenic through several in silico analyses using SIFT , PolyPhen2 , and MUpro programs . Most of all , a recent report on the same mutation ( p . I43N ) also strongly suggested that the mutation is the genetic cause of this family [10] . The clinical manifestations are summarized in Table 2 . The age at onset ranged from 6 to 18 years among the three affected individuals . In all patients , distal leg muscle weakness , atrophy , and frequent falling were first noticed . They also exhibited bilateral hand muscle weakness that caused difficulty in writing , cooking , and fine finger control . Their hand atrophy was first noticed in the dorsal interosseous and thenar muscles , but their hypothenar muscle bulk and strength was relatively preserved . Neurological examination indicated that muscle weakness and atrophy started and predominated in the distal portions of the legs , and were noted to a lesser extent distally in the upper limbs . All of patients had pes cavus without scoliosis . Foot dorsiflexion ( MRC , G0/5 to G2/5 ) was markedly weaker than foot plantarflexion ( G4+/5 to G5/5 ) and finger abduction ( G4/5 to G4+/5 ) . Among the three patients , patient III-1 exhibited the most severity in muscle disability in the upper and lower limbs . All of them were able to walk independently with or without ankle-foot orthosis . Vibration sense was more severely affected than pain sense . Areflexia was noticed in the early stages of the disease , but pathologic reflexes were not found . On neurological and electrophysiological examination , her mother ( I-2 ) showed normal findings . Based on history taking , her father ( I-3 ) experienced no difficulty with walking . SNAPs of the sural nerve were not detected at the early stage of the disease . Although CMAPs of the tibial motor nerves were evoked in all patients , SNAPs of sural nerves were not elicited at all ( Table 3 ) . All three patients had reduced amplitudes for the peroneal innervated foot muscles , and those responses were lesser than the responses from the thenar muscles . Needle EMG showed scattered fibrillation potentials and neurogenic motor unit action potentials . MRIs of the three patients revealed predominant signal changes in the anterior and lateral compartments of the lower leg muscles ( Fig 2 ) . T1-weighted coronal images showed more muscle atrophy and fatty replacement in the calf muscles than in the hip and thigh muscles ( S1 Fig ) . We could observe a sequential pattern of muscle involvement associated with disease duration ( Fig 2 ) . It is noteworthy that the lateral compartment muscles including the peroneus longus and brevis were initially involved and they showed the most severe fatty hyperintense signal changes . In the early stage of the disease , diffuse fatty infiltration was observed in the lateral compartment muscles , whereas it was less severe in the anterior compartment muscles including the tibialis anterior , extensor hallucis longus , and extensor digitorum longus ( Fig 2 ) . In later stages , the posterior compartment muscles including the soleus , gastrocnemius , and tibialis posterior in the calf were still relatively unaffected ( Fig 2 ) . When we performed follow-up lower leg MRI studies at a 5-year interval , we could observe disease progression ( S1 Fig ) . Muscle atrophy and fatty involvement in the calf were more rapid and severe than those in the thigh and hip muscles . At the hip and thigh level , the axial MR images were almost normal except for the bilateral semitendinosus and vastus lateralis muscles . At the calf level , the anterior and lateral compartment muscles showed more prominent fatty replacement than the posterior compartment muscles . In addition , brain MRI showed normal findings in the three patients ( S2 Fig ) . Light microscopic examination of longitudinal sections and cross sections of nerve fibers showed a slight decrease in the size of nerve fascicles as well as a decrease in the number of large myelinated fibers ( MFs ) , and endoneurial fibrosis . Semi-thin transverse sections showed a normal range of MFs ( 8 , 601/mm2 ) with loss of large MFs , myelin abnormalities , onion bulbs , and regenerating axonal clusters ( Fig 3 ) . The histogram showed a unimodal distribution pattern of MFs . The range and average diameter of MFs were 0 . 86–12 . 61 μm and 2 . 59 μm , respectively . After excluding the largest MF ( marked with an asterisk at the right upper corner of Fig 3A , diameter: 12 . 61 μm ) , the range and average diameter of the remaining MFs were 0 . 86–7 . 04 μm and 2 . 57 μm , respectively . The range and average diameter of MFs in the distal sural nerve of a normal 45-year-old female are 1 . 8–14 . 8 μm and 5 . 2 μm , respectively . The MF% area in this case was 5 . 5% ( distal sural nerve of a normal 45-year-old female: 26 . 9% ) . The range and average of the g-ratio were 0 . 33–0 . 86 and 0 . 62±0 . 09 , respectively ( mean g-ratio in the age group of 21–50 years: 0 . 66 ) . g-ratios above 0 . 7 comprised 18 . 7% of MFs and g-ratios less than 0 . 4 comprised 0 . 6% of the MFs . Electron microscopic examination showed frequent onion bulb formation , large or medium sized MFs with myelin abnormalities , regenerating axonal clusters , and thin or thick MFs ( Fig 3 ) . To investigate the PMP2 mutation-associated pathogenesis , we generated transgenic mouse models for both wild type and mutant PMP2 . Except for the peripheral neuropathy , both mice exhibited normal phenotype including normal development and breeding . To determine whether mice overexpressing human PMP2 exhibit motor deficits , a tail suspension test was performed . Both PMP2-WT and PMP2-I43N transgenic mice showed hind limb folding into the abdomen with infrequent clasping of their hind limbs ( Fig 4 ) . Both transgenic mice also showed reduced motor performance on the rotarod test . Therefore , transgenic mice overexpressing either wild type or mutant PMP2 definitely exhibited motor deficits compared to control mice . To determine the nerve integrity in the hind limbs , nerve conduction was examined on both sides of the sciatic nerve at 5 months of age . Both transgenic mice showed significantly reduced MNCV compared to control mice ( Fig 4 ) . However , there was no difference in CMAP between control and PMP2 transgenic mice . These data suggest that the phenotypes of both PMP2 transgenic mice were similar to CMT1 rather than CMT2 , which is caused by axonal defect and shows reduced CMAP . Semi-thin sections of the sciatic nerve displayed a reduced number of large myelinated fibers in PMP2-I43N mice compared to control mice , while PMP2-WT mice showed a pattern similar to the control mice . The average g-ratios were 0 . 75 ± 0 . 07 ( control ) , 0 . 71 ± 0 . 09 ( PMP2-WT ) , and 0 . 69 ± 0 . 09 ( PMP2-I43N ) . Electron microscopic analysis revealed mixed forms of demyelinated or dysmyelinated fibers in both PMP2-WT and PMP2-I43N mice compared to control mice ( Fig 5 ) . When we analyzed the teased nerve , the PMP2 transgenic mice revealed more frequent occurrence of Schmidt-Lanterman incisures ( SLIs ) ( Fig 6 ) . In addition , the mean internodal length in PMP2-WT and PMP2-I43N mice was 259 . 0 ± 14 . 6 μm and 229 . 3 ± 10 . 9 μm , which was significantly shorter than that in healthy controls ( 442 . 4 ± 15 . 8 μm; P < 0 . 001 ) ( Fig 6 ) . However , the difference in internodal length between PMP2-WT and PMP2-I43N mice was not statistically significant . Considering the shortened internodal length in the transgenic mice , the number of SLIs per myelinating Schwann cell was within a similar range of 9 . 6 ( PMP2-I43N ) to 13 . 1 ( PMP2-WT ) per cell . Collectively , these data suggest that the overexpression of either wild type or mutant PMP2 affects Schwann cell integrity , resulting in reduced internodal length as well as demyelination , which might cause a demyelinating phenotype in transgenic mice .
This study strongly suggests that PMP2 mutation is the underlying cause of an autosomal dominant CMT1 phenotype . The same mutation in PMP2 was also suggested as a strong candidate for the cause of another CMT1 family [10] . Moreover , the pathogenicity of PMP2 mutation had been demonstrated in a zebrafish model and the relevance of the CMT1 phenotype was also observed in our transgenic mouse models . Exome data revealed many functionally significant variants in the CMT-relevant genes from the affected individuals . Although they were not considered as the underlying cause of the CMT1 phenotype , some of them may affect clinical severity as the modifying factor ( s ) . The clinical features of the present family were similar to PMP22 duplication . The age at onset among the patients was the first and second decades . The nerve conduction velocities in all three patients were below 38 m/s . Muscle weakness and atrophy started and predominated in the distal portions of the legs , and were noted to a lesser extent distally in the upper limbs . Moreover , the histopathological findings including onion bulb formation were consistent with the demyelinating type . In all patients , SNAPs of the sural nerve were not evoked at the early stage of the disease . Additionally , there were clinical heterogeneities within the present family . The ages at onset of gait disturbance were in the first decade among the affected sons ( III-1 and III-3 ) , and in the second decade for their mother ( II-4 ) . Hand muscle weakness appeared in the third decade among the affected sons ( III-1 and III-3 ) , whereas it appeared in the fifth decade in their mother ( II-4 ) . Furthermore , the results of nerve conduction studies were mild in the mother , but were more severe in her sons . MRI analysis revealed a distinct pattern of muscular involvement in patients with PMP2 mutation . Marked fatty replacement was observed in the calf muscle compared to the thigh muscle , which was consistent with the hypothesis of length-dependent degeneration of motor axons . These results are useful for discerning motor neuron disease from peripheral neuropathies like CMT1 with PMP2 mutation . In addition , we could observe a sequential pattern of muscle involvement associated with disease duration . In the early disease stage , the peronei muscles were involved , and they showed the most severe fatty replacement . The tibialis anterior muscles were also involved to a lesser degree . In the later stage , the soleus muscle in the calf was found to be relatively spared . These findings were similar to the demyelinating MR images of CMT1A patients with PMP22 duplication , and they differed from the axonal MR images of CMT2A patients with MFN2 mutations . Compared to CMT2A , in which severe and predominant involvement of the soleus muscles has been reported , the present patients showed relative sparing of the soleus muscles , which suggests that the etiologies and pathophysiologies of CMT2A with MFN2 mutation are different . Therefore , MRIs of patients with PMP2 mutation were well related to the demyelinating CMT neuropathy . PMP2 is a peripheral myelin protein gene that encodes a small basic P2 protein . P2 is also called M-FABP/FABP8 and is a member of a family of fatty acid binding proteins with maximal specificity of expression in the peripheral myelin [11–13] . P2 enhances myelin membrane stability and lipid dynamics [14] . P2 forms a β-barrel , which is well conserved through the species as well as throughout the FABP family subtypes with 10 anti-parallel β-strands and an N-terminal helix-loop-helix cap , where fatty acids are bound inside the barrel [15–17] . Molecular docking simulations show that the best candidate for the P2 pocket is cholesterol , one of the most abundant lipids in myelin , with a polar head that suits the electrostatic characteristics of the binding site [16] . Hydrophobic contacts surround the cavity entrance while hydrophilic contacts are located deep in the binding pocket . The mutation site ( p . I43N ) lies inside the pocket and substitutes a non-polar Ile with a polar Asn . Therefore , the substitution may possibly alter the dynamics of the fatty acid capacitation . In silico analyses also predict that PMP2 mutation might affect protein function and stability . The demyelinating features were also observed in PMP2 transgenic mouse models . Both PMP2-WT and PMP2-I43N transgenic mice exhibited reduced MNCV , which is compatible with the CMT1 phenotype . We found that PMP2 transgenic mice have shorter internodal length . Since internodal length is considered to affect the velocity of nerve impulse conduction , lower MNCV in transgenic mice might be due to the shorter internodal length [18] . Previously , shortened internodal length was reported in CMT1A patients and a Periaxin null mouse model [19 , 20] . In addition , a cell culture model using Trembler J mouse also showed reduced internodal length [21] . Therefore , shortened internodal length might be a characteristic feature of CMT1 . Since both transgenic mice either overexpressing wild type or mutant PMP2 exhibited a similar neuropathic phenotype , we could not determine whether p . I43N mutation was pathogenic or not until the zebrafish model for PMP2 was reported [10] . Overexpression of either wild type or mutant ( p . I43N ) PMP2 in zebrafish also led to defects in axonal branching . Therefore , these data imply that PMP2 might result in peripheral neuropathy in the same manner as the well-characterized PMP22 gene , which affects myelination by either mutation in one allele or overdose of the wild type gene . In addition , in vitro assay showed that overexpression of both wild type and mutant PMP2 in rat Schwann cell induced the markers of endoplasmic reticulum stress , which might result in cell death ( S3 Fig ) . Compared to the zebrafish model , our mouse models exhibited a clear CMT1 phenotype . The zebrafish model showed that aberration in PMP2 expression exhibited motor neuron phenotypes such as axonal branching , which is somewhat different from the primary pathogenesis of CMT1 . On the other hand , the zebrafish model clearly demonstrated the pathogenicity of p . I43N mutation . Overexpression of PMP2-I43N in zebrafish showed limited benefits compared to wild type PMP2 . In this study , we could not directly observe the PMP2 mutation-associated neuropathy . Phenotypic severity in the PMP2-I43N transgenic mouse compared to the PMP2-WT mouse was limited . Although PMP2-I43N transgenic mice exhibited worse parameters including rotarod performance , MNCV , g-ratio and internodal length than PMP2-WT mice , the differences were not statistically significant . For clear demonstration of PMP2 mutation-associated neuropathy , further studies using knock-in mouse models are needed . In this study , we first report the clinical features of patients with CMT1 caused by PMP2 mutation as well as the characterization of transgenic mice . This report might expand the genetic and clinical features of CMT and a further mechanism study will enhance our understanding of the PMP2 associated peripheral neuropathy .
The study involving human patients was approved by the Institutional Review Board of Sungkyunkwan University , Samsung Medical Center ( approval number: 2013-10-066 ) . Written informed consent was obtained from all participants . All animal studies were conducted according to protocols approved by the Institutional Animal Care and Use Committees of Samsung Medical Center ( approval number: 2013-080-2002 ) . This study enrolled a Korean autosomal dominant CMT1 family ( family ID: FC183 , Fig 1 ) with 9 members ( 3 affected and 6 unaffected members ) who showed no 17p12 duplication/deletion . In addition , 500 healthy controls , who had no family history of neuromuscular disorders , were enrolled after careful clinical and electrophysiological examination . Patients were evaluated through a detailed history including the assessment of motor impairments , sensory loss , deep tendon reflexes , and muscle atrophy . Age at onset was determined by asking the patients their ages , when symptoms such as distal muscle weakness , foot deformity , or sensory change first appeared . Muscle strengths of flexor and extensor muscles were assessed manually using the standard Medical Research Council ( MRC ) scale [22] . In order to determine physical disability , we used two scales , a functional disability scale ( FDS ) and a CMT neuropathy score ( CMTNS ) [23 , 24] . Sensory impairments were assessed in terms of the level and severity of pain , temperature , vibration and position , then pain and vibration senses were compared . The Nine-Hole Peg Test ( 9-HPT ) was performed using both dominant and non-dominant hands; five consecutive trials for the dominant hand , followed immediately by another five consecutive trials for the non-dominant hand . Electrophysiological studies were carried out in the 3 affected individuals . Nerve conduction studies were performed by placing surface electrodes on median , ulnar , peroneal , tibial , and sural nerves . MNCVs for the median and ulnar nerves were determined by stimulating at the elbow and wrist while recording compound muscle action potentials ( CMAPs ) over the abductor pollicis brevis and adductor digiti quinti , respectively . In the same manner , the MNCVs of the peroneal and tibial nerves were determined by stimulating at the knee and ankle , while recording CMAPs over the extensor digitorum brevis and adductor hallucis , respectively . Sensory nerve conduction velocities ( SNCVs ) were obtained over a finger-wrist segment from the median and ulnar nerves by orthodromic scoring , and were also recorded for the sural nerves . Electromyography ( EMG ) was performed for the first dorsal interosseous , biceps brachii , tibialis anterior , medial gastrocnemius , and vastus lateralis muscles . The three patients ( Fig 1: II-2 , III-2 , and III-4 ) underwent examination by lower limb MRIs of the hip , thigh , and calf muscles . MRI was performed by using a 1 . 5-T system ( Siemens Vision , Siemens , Germany ) . The imaging was conducted in the axial [field of view ( FOV ) 24–32 cm , slice thickness 10 mm , and slice gap 0 . 5–1 . 0 mm] and coronal planes ( FOV , 38–40 cm; slice thickness 4–5 mm; and slice gap 0 . 5–1 . 0 mm ) . The following protocol was used: T1-weighted spin-echo ( SE ) ( TR/TE 570-650/14-20 ms ) , T2-weighted SE ( TR/TE 2800-4000/96-99 ms ) , and fat-suppressed T2-weighted SE ( TR/TE 3090-4900/85-99 ms ) . In addition , brain MRI was performed in the three patients ( II-4 , III-1 , and III-3 ) by using a 1 . 5-T system ( Siemens Vision , Siemens , Germany ) . Whole brains were scanned using a slice thickness of 7 mm and a 2-mm interslice gap to produce 16 axial images . The imaging protocol consisted of T2-weighted spin echo ( TR/TE = 4 , 700/120 ms ) , T1-weighted spin echo ( TR/TE = 550/12 ms ) , and fluid-attenuated inversion recovery ( FLAIR ) ( TR/TE = 9 , 000/119 ms , inversion time 2 , 609 ms ) images . Patient II-4 underwent distal sural nerve biopsy at the age of 56 years . Pathological examinations included light and electron microscopic analysis . One sural nerve fragment was fixed in 10% formalin , embedded in paraffin , and stained with hematoxylin-eosin , modified Masson’s trichrome , Luxol fast blue , and Bodian stain . For the electron microscopic observation , another fragment was immediately fixed with 2% glutaraldehyde in 0 . 025 M cacodylate buffer at pH 7 . 4 and post-fixed in 1% osmium tetroxide . Epon-embedded semi-thin and ultra-thin sections were prepared for light and ultra-structural examinations . DNA was purified from peripheral blood using the QIAamp blood DNA purification kit ( Qiagen , Hilden , Germany ) . Exome sequencing was performed on 5 members of the FC183 family ( 3 affecteds: II-4 , III-1 , III-3; and 2 unaffecteds: II-1 , II-3 ) . Exomes were captured using the SeqCap EZ ver 3 . 0 ( Roche-NimbleGen , Madison , WI ) , and sequencing was performed using the HiSeq 2000 Genome analyzer ( Illumina , San Diego , CA ) . Sequences were mapped to the human reference genome UCSC assembly hg19 , and variants were called by the SAMtools program ( http://samtools . sourceforge . net/ ) . We first selected functionally significant variants ( missense , nonsense , exonic indel and splicing site variants ) and then filtered out polymorphic variants registered in the dbSNP ( http://www . ncbi . nlm . nih . gov ) , 1000 Genomes project database ( 1000G , http://www . 1000genomes . org/ ) , Exome Sequencing Project ( ESP , http://evs . gs . washington . edu/EVS/ ) , and Exome Aggregation Consortium ( ExAC , ( http://exac . broadinstitute . org/ ) . Next , we further selected variants that cosegregated with the three affected individuals within the family and were not found in the 500 healthy controls . Conservation analysis of protein sequences was conducted with MEGA ver 5 . 05 program [25] . Genomic evolutionary rate profiling ( GERP ) scores were determined with the GERP++ program ( http://mendel . stanford . edu/SidowLab/downloads/gerp/index . html ) . In silico analyses were performed to predict the deleterious nature of the protein function due to amino acid substitution using SIFT ( http://sift . jcvi . org/ ) and Polyphen-2 ( http://genetics . bwh . harvard . edu/pph2/ ) ; while protein stability distortion due to amino acid substitution was predicted with the MUpro program ( http://mupro . proteomics . ics . uci . edu/ ) . To obtain the human PMP2 gene , total mRNA from HEK 293 cells was used as a template for cDNA synthesis and PCR amplification . Mutation ( p . I43N ) of PMP2 was generated by site-directed mutagenesis using the QuikChange Site-Directed Mutagenesis kit ( Stratagene , La Jolla , CA ) and all sequences were confirmed by capillary sequencing . To establish transgenic mouse models for wild type and mutant PMP2 , cloned vectors were injected into fertilized eggs . The eggs were implanted into surrogate female mice , then transgenic mice were generated . To evaluate the motor coordination of PMP2 transgenic mice , rotarod tests were performed on a 3 cm horizontal rotating rod at a speed of 2 m/min . To adapt to the test , mice were pre-trained for one week . For the electrophysiological test , 10 of each control and transgenic mice , aged 5 months and weighing 25–30 g , were used . The mice were anesthetized with Zoletil ( 50 mg/kg ) intraperitoneally ( Virbac , Seoul , Korea ) and the fur from the distal back and the hind limbs was completely removed . The CMAP and MNCV were determined by using the Nicolet VikingQuest ( Natus Medical , San Carlos , CA ) as previously described [26] . Histological analysis of mouse sciatic nerves was performed in the same manner with the patient samples . To visualize the Schmidt-Lanterman incisures ( SLI ) and nucleus , teased nerves were stained with rhodamine conjugated Phalloidin ( Thermo Fisher Scientific , Waltham , MA ) and mounted with 4′ , 6′-diamidine-2′-phenylindole dihydrochloride ( DAPI ) containing Vectashield ( Vector Laboratories , Burlingame , CA ) as previously reported [27] . Internodal length was determined by measuring the distance of each nuclei stained with DAPI according to a previous report [28] . All animals were studied with a blind test . Comparison between normal and transgenic mice were made by Student’s t-test . P<0 . 05 was considered statistically significant . | Isolation of causative mutation is still challenging in genetic diseases with a variety of genetic causes . We discovered a mutation in a novel gene from a family exhibiting a peripheral neuropathy by virtue of next-generation sequencing . Although the family shows characteristic clinical features of hereditary motor and sensory neuropathy , we could not find a mutation from well-known genes . To demonstrate the clinical relevance of the novel gene , we generated transgenic mice , which carry the patients’ mutation within their chromosome . The transgenic mice exhibited the same phenotype as the patients including peripheral neuropathic symptoms and reduced locomotor function . We also observed the affected peripheral nervous system through electron microscopy . It seems that the expression of the mutant protein impairs the myelin of peripheral nervous system . These data might expand the genetic , clinical , and pathophysiological features of the peripheral neuropathy and a further investigation will enhance our understanding of disease in the peripheral nervous system . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"medicine",
"and",
"health",
"sciences",
"legs",
"diagnostic",
"radiology",
"charcot-marie-tooth",
"disease",
"genetic",
"diseases",
"limbs",
"(anatomy)",
"magnetic",
"resonance",
"imaging",
"animal",
"models",
"mutation",
"model",
"organisms",
"bioassays",
"and",
"physiological",
"analysis",
"muscle",
"electrophysiology",
"nerve",
"conduction",
"study",
"research",
"and",
"analysis",
"methods",
"neuropathy",
"musculoskeletal",
"system",
"imaging",
"techniques",
"clinical",
"neurophysiology",
"mouse",
"models",
"electrophysiological",
"techniques",
"clinical",
"genetics",
"point",
"mutation",
"radiology",
"and",
"imaging",
"diagnostic",
"medicine",
"peripheral",
"neuropathy",
"anatomy",
"neurology",
"genetics",
"biology",
"and",
"life",
"sciences"
]
| 2016 | A Mutation in PMP2 Causes Dominant Demyelinating Charcot-Marie-Tooth Neuropathy |
The prevalence of antibiotic resistance genes in pathogenic bacteria is a major challenge to treating many infectious diseases . The spread of these genes is driven by the strong selection imposed by the use of antibacterial drugs . However , in the absence of drug selection , antibiotic resistance genes impose a fitness cost , which can be ameliorated by compensatory mutations . In Streptococcus pneumoniae , β-lactam resistance is caused by mutations in three penicillin-binding proteins , PBP1a , PBP2x , and PBP2b , all of which are implicated in cell wall synthesis and the cell division cycle . We found that the fitness cost and cell division defects conferred by pbp2b mutations ( as determined by fitness competitive assays in vitro and in vivo and fluorescence microscopy ) were fully compensated by the acquisition of pbp2x and pbp1a mutations , apparently by means of an increased stability and a consequent mislocalization of these protein mutants . Thus , these compensatory combinations of pbp mutant alleles resulted in an increase in the level and spectrum of β-lactam resistance . This report describes a direct correlation between antibiotic resistance increase and fitness cost compensation , both caused by the same gene mutations acquired by horizontal transfer . The clinical origin of the pbp mutations suggests that this intergenic compensatory process is involved in the persistence of β-lactam resistance among circulating strains . We propose that this compensatory mechanism is relevant for β-lactam resistance evolution in Streptococcus pneumoniae .
Streptococcus pneumoniae is the causal agent of human infections such as otitis , pneumonia and meningitis , which particularly affect pediatric patients . Penicillin and other β-lactams ( βLs ) are the treatments of choice for any pneumococcal infection . However , β-lactam resistance has been growing steadily since the first clinical resistant isolate was reported in the 1960s , with its dissemination greatly threatening the clinical efficacy of these compounds [1] . βLs act by binding to the active sites of PBPs , which are involved in peptidoglycan synthesis , thus altering the normal cell wall formation and inducing cell death by lysis . Specifically , PBPs catalyze the last two stages of peptidoglycan biosynthesis , namely transglycosylation and transpeptidation . Its ability to uptake exogenous DNA by induction of competence allows S . pneumoniae to acquire antibiotic resistance mutations . β-lactam resistance results from the acquisition of mutations within the pbp2b , pbp2x , and pbp1a genes , causing multiple amino acid changes in PBP2b , PBP2x , and PBP1a , which decrease the enzyme affinity for βLs [2] . Exogenous DNA containing pbp mutations can be provided by β-lactam resistant ( βLR ) strains of S . pneumoniae or by other Streptococcus species that cohabit the same niche , and is incorporated into the chromosome by homologous recombination [3] . The rise in βL resistance detected among circulating strains is due to their ability to acquire pbp genes , as mentioned above , in association with the wide therapeutic use of βL compounds , which could act as a selective pressure [4] . In the absence of antibiotic therapy , bacterial antibiotic resistance genes commonly result in reduced fitness . Nevertheless , by acquiring intra or intergenic compensatory mutations , certain clinical and laboratory strains are able to restore this fitness cost [5] , [6] . Presumably , this compensation allows these strains to compete with susceptible strains in their natural niches . In S . pneumoniae , ciprofloxacin-resistant isolates have a low prevalence due to a fitness cost imposed by point mutations in the genes that encode DNA polymerase and DNA gyrase [7] , [8] . A putative relationship has also been suggested between βL resistance and the loss of virulence in pneumococcal clinical strains [9] , [10] . Likewise , Rieux and coworkers [11] described a loss of virulence in single and double isogenic βLR mutants , obtained by the transformation of pbp2x and pbp2b mutations amplified from βLR clinical isolates . Related to this , Trzcinski and coworkers , studying clinical isolates [12] , reported an increase in fitness cost when βL resistance was conferred by the pbp2b pbp2x pbp1a mutations acquired by sequential transformation events . In this work , we studied the clinical βLR strains belonging to a new serotype-14 variant of the Spain9V-3 international clone . This variant was isolated in Argentina and is indistinguishable by MLST and sequence analysis of the cps genes from those isolated in Baltimore , USA [13] . We demonstrated that the sequential acquisition of pbp mutations to develop βL resistance is closely associated to an intergenic compensatory process , which can restores the fitness cost imposed by pbp2b mutations and may favor the persistence and spread of βL resistance in S . pneumoniae .
The fitness cost imposed by pbp2b , pbp2x and pbp1a mutations that confer βL resistance in S . pneumoniae was evaluated . These pbp alleles were obtained from βLR clinical strains that belonged to a new serotype-14 variant of the Spain9V-3 international clone , identified and characterized in our laboratory [13] . To generate strains bearing alleles of pbp2b , pbp2x , and pbp1a from clinical strains that conferred resistance , we used PCR to amplify the DNA region that encodes the transpeptidase domain of each mutated pbp gene ( in pbp1a , the glycosyltransferase domain was also included; Fig . S1 in Supporting Information S1 ) , where the βL resistance mutations are commonly localized . These PCR products containing the pbp mutations were introduced by transformation into the βL-susceptible laboratory strain Cp1015 , and βLR mutants were selected by growth on agar plates supplemented with piperacillin for pbp2b , or with cefotaxime for pbp2x and pbp1a ( Table S1 in Supporting Information S2 ) . The pbp mutations were integrated into the chromosome by homologous recombination . Although no growth alterations for single pbp2x or pbp1a mutants were found , which presented similar growth curves to that of Cp1015 , pbp2b mutants revealed an uncommonly prolonged lag-phase ( Fig . S2 in Supporting Information S1 ) . Then , due to the fact that clinical strains ( from which pbp genes were amplified to generate single pbp mutants ) grew as well as the wild-type strain ( data not shown ) , we decided to investigate whether the fitness cost resulting from pbp2b mutations was compensated for by associating the pbp2x and pbp1a genes amplified from the same βLR strain donor of pbp2b . For this purpose , double and triple pbp mutants were obtained by sequential transformation of single and double pbp mutants ( in that order ) , and the βLR mutant selection was made by using the MIC ( Minimal Inhibitory Concentration ) increment produced by the acquisition of different pbp mutations ( Table 1 ) . For example , a 14-fold increase occurred in the piperacillin MIC level for the pbp2b mutant; a 10-fold increase was found in the cefotaxime MIC level for the pbp2x or pbp1a mutants ( single or double pbp mutants ) , and a 14-fold increase resulted in the cefotaxime MIC level for triple pbp mutants . The individual and double pbp mutants displayed a transformation rate higher than 10−2/µg of DNA , similar to the wild-type strain . Natural competence was also assayed , but no alterations in spontaneous transformability were detected ( Fig . S3 in Supporting Information S1 ) . We observed a partial restitution of the growth curve for the double pbp2b pbp2x and pbp2b pbp1a mutants , while a clear restoration was found for the triple pbp mutant ( Fig . S2 in Supporting Information S1 ) . Interestingly , all seven pbp2b mutants constructed by transformation with different mutated pbp2b genes obtained from other non-susceptible clinical isolates ( Cba strains , Table S1 in Supporting Information S2 ) and the Spain9V-3 ATCC strain displayed similar growth alterations ( data not shown ) . These pbp2b alleles were sequenced , and a comparison between them revealed that the DNA sequences obtained from Cba-19 , Cba-28 , Cba-52 , Cba-62 and ATCC 700671 ( clone Spain9V-3 ) were identical ( Fig . S4 in Supporting Information S1 ) . It was also found that 5 amino acids were conserved in the transpeptidase domain for all the pbp2b sequences ( P417 , E443 , I460 , G481 and A494 ) obtained from the different βLR clinical strains . These results suggest that these substitutions could be responsible for the fitness cost detected in the pbp2b mutants . The pbp2b sequences of mutants constructed in the Cp1015 strain were identical to those obtained from the clinical strains , with each pbp2b sequence being coincident with the original pbp2b from the clinical strains used to transform the wild-type strain . Hereafter , the strain reference will be expressed as superscripts in the pbp mutations; for example , the pbp2b mutation from strain Cba-28 that was transformed into the Cp1015 strain will be indicated as pbp2b28 . All findings about compensation of growth alterations were similar to the results obtained by competitive fitness assays in vitro , when comparing the βL-susceptible Cp1015 strain with isogenic pbp mutants . The compensatory effect among pbp mutations was evident not only in triple , but also in pbp2b19 pbp1a19 , pbp2b19 pbp2x19 , pbp2b28 pbp1a28 , and pbp2b9V3 pbp2x9V3 double mutants , recovering their values of relative fitness of 0 . 95–1 . 10 in the case of the triple pbp mutants ( Table 1 ) . We also observed that single pbp1a mutations obtained from Cba-19 and Cba-28 increased fitness in the wild-type strains ( Table 1 ) , and therefore should be considered as increasing-fitness mutations . Interestingly , we found a direct correlation between fitness compensation and an increase in the spectrum and level of βL resistance ( Fig . 1; Table 1 ) . The acquisition of pbp2x and pbp1a mutations from clinical βLR strains other than Cba-28 was also able to compensate the fitness cost found in pbp2b28 mutants . We observed that fitness cost was ameliorated in 7 of the 8 double pbp2b28 pbp2x mutants analyzed , which were constructed by transforming pbp2x genes from donor βLR strains different from Cba-28 into the pbp2b28 mutant ( Table S2 in Supporting Information S2 ) . When the transpeptidase regions of pbp2x sequences from the different mutants constructed in Cp1015 were compared , we observed that the pbp2x sequences in 4 mutants ( pbp2b28 pbp2x12 , pbp2b28 pbp2x52 , pbp2b28 pbp2x54 and pbp2b28 pbp2x62 ) presented 10 conserved substitutions compared with Cp1015 , with the exception of pbp2x52 , which showed D278 in place of N278 ( Fig . S5 in Supporting Information S1 ) . Coincidently , pbp2x52 was the only one of these 4 pbp2x alleles that could not ameliorate the fitness cost of pbp2b28 ( Table S2 in Supporting Information S2 ) . This result suggests that the N278 residue might be involved in the compensatory mechanism of fitness , at least for the set of mutations present in pbp2x12 , pbp2x54 and pbp2x62 . Each pbp2x sequence used to transform Cp1015 was identical to the original pbp2x obtained from clinical strains ( data not shown ) . To corroborate the fitness compensation of pbp mutations observed in vitro , we also evaluated the relative fitness in vivo of single , double and triple pbp mutants using a model of intraperitoneal infection in C57BL/6 mice . Due to the fact that Cp1015 is an avirulent strain , we transferred the pbp mutations into the D39 strain , a virulent capsulated strain , which is able to infect C57BL/6 mice [14] . Single pbp mutants were obtained in D39 as described for Cp1015 . Curiously , however , only crossed double and triple pbp mutants were recovered by transformation of the pbp2x and pbp1a genes from the Cba-19 strain into the D39 pbp2b9V3 mutant ( Table 1 ) , indicating that the βL resistance conferred by these pbp mutations was dependent on the genetic background of the recipient strain . In agreement with the data obtained by assays in vitro , we observed a similar compensatory phenomenon for the D39 pbp mutants in a mice model , with values of relative fitness of 0 . 27 for pbp2b9V3 , 0 . 80 for pbp2b9V3 pbp2x19 , and 1 . 14 for the triple pbp2b9V3 pbp2x19 pbp1a19 mutant being obtained ( Table 1 ) . βL resistance is complex and it has been reported that is conferred by mosaic DNA fragments containing several mutations in the transpeptidase domain , which decrease the binding of βLs . With the purpose of identifying the amino acid substitutions that could be involved in the fitness compensation mechanism described , sub-parts of the pbp2x and pbp1a genes from the βLR Cba-28 strain were amplified by PCR ( see schemes shown in Figs . S6 and S7 in Supporting Information S1 ) , and PCR products were independently transformed into the pbp2b28 mutant , with the transformants being selected by cefotaxime resistance . We showed that fragments A2 , A3 ( both amplified from pbp1a28 ) and X4 ( amplified from pbp2x28 ) conferred resistance , but only A2 could also compensate the fitness alterations ( Table S3 in Supporting Information S2 ) , reaching the same level as that obtained when pbp2b28 was transformed with the pbp1a28 fragment containing the glycosyltransferase and transpeptidase domains ( Table 1 ) . We analyzed the amino acid sequences deduced from the DNA sequences of A2 and A3 , and found four substitutions shared between both regions ( Fig . S8 in Supporting Information S1 ) , which are probably involved in βL resistance . Interestingly , the E285Q mutation , localized in the transpeptidase domain [15] , was only present in A2 ( from pbp1a28 ) . We propose that the E285Q mutation may have contributed to the fitness compensation mechanism . The presence of all these mutations was found in the double mutants pbp2b28 pbp1a28 and pbp2b28 pbp2x28 constructed in the Cp1015 genetic background as well as in the clinical Cba-28 strain . In contrast with our results , Trzcinski and coworkers [12] reported that the fitness cost of pneumococcal penicillin-resistant strains increased with the number of resistant pbp alleles acquired . On comparing the predicted amino acid sequences of the pbp alleles studied here with those obtained by Trzcinski and coworkers , we found clear differences between them , detecting 9 , 11 and 12 substitutions in PBP2x , PBP1a and PBP2b , respectively ( Figs . S9 , S10 and S11 in Supporting Information S1 ) . This suggests that the discrepancy found between these works may be due to specific pbp mutations . It is known that PBPs are responsible for peptidoglycan synthesis . These proteins form the main component of the cell wall , and are involved in longitudinal wall growth and the cell division process . Morlot and coworkers [16] have proposed certain functions for PBP2b , PBP2a and PBP1a in S . pneumoniae , but the specific roles of PBPs in the growth and division phases are not yet well understood in bacteria . PBPs are also recognized as the targets for βLs , the most frequently used antimicrobials for treating bacterial infections . Therefore , considering that the interaction between drug resistance and the cell division processes remains unclear [4] , we decided to investigate this connection . First , we examined the effects of pbp mutations on S . pneumoniae cell morphology , and then the putative cell division alterations . Microscopic examinations of single pbp1a28 and pbp2x28 mutants in Cp1015 strain did not display any morphological alterations , whereas the pbp2b28 mutant ( Fig . 2A ) and all 10 strains bearing pbp2b mutations from clinical strains showed a rod shape ( data not shown ) . This atypical phenotype was also found for pbp2b mutants with a different genetic background to that of Cp1015 , such as the R6 and D39 laboratory strains ( data not shown ) . Curiously , 46% of pbp2b28 cells had a rod-like shape ( Fig . 2A ) , with the rest showing an apparent wild-type morphology . Electron microscopy revealed that the coccoid-shaped cells exhibited a variety of cell wall defects , including an abnormal septum position , atypical intracellular structures and frequent asymmetrical divisions ( Figs . 3 and S13 in Supporting Information S1 ) compared with the wild-type strain ( Figs . 3 and S12 in Supporting Information S1 ) , with the rod-shaped cells exhibiting multiple septa ( Figs . 3 and S13 in Supporting Information S1 ) . Moreover , the double pbp2b28 pbp2x28 mutant displayed a coccoid morphology but with atypical septum localizations and peptidoglycan accumulation ( Figs . 3 and S14 in Supporting Information S1 ) . However , the triple pbp mutants had wild-type cell morphology and showed no ultrastructural alterations ( Figs . 3 and S15 in Supporting Information S1 ) . When cells were stained with fluorescein-labeled vancomycin ( Van-FL ) , which localizes to sites of nascent peptidoglycan synthesis and clearly marks the septum location in the wild-type strain [17] ( Fig . S16 in Supporting Information S1 ) , an abnormal septum pattern was revealed in rod-shaped cells of the pbp2b28 mutants , suggesting a clear alteration in cell division ( Fig . 4 ) . However , this phenomenon was compensated in the double ( pbp2b28 pbp2x28 or pbp2b28 pbp1a28 ) and triple pbp mutants ( Fig . 4 ) . The morphological variation was confirmed by flow cytometry analysis , which allowed determining the population distribution of pneumococci by cell size . These assays showed a displacement favoring a larger cell size in the pbp2b28 mutant , and the restoration of normal size in the triple pbp mutant ( Fig . 2B ) . These results suggest that the septal alterations and the cellular enlargement found in the pbp2b28 mutant could have been responsible for its growth retardation . To investigate the putative cause of these effects on cell morphology and fitness , we analyzed the stability of the proteins encoded by these pbp mutated genes using immunoblotting and inhibiting the protein synthesis by the addition of kanamycin . For these assays , we constructed C-terminal HA tagged PBPs to permit the detection of the proteins using an anti-HA monoclonal antibody . The gene constructs were inserted into the chromosome by insertion-duplication as described previously [18] , and these genes were expressed as single copies under the control of their native promoters . These pbp-HA mutants showed the same phenotypes as the original pbp mutants , as well as the compensatory effects demonstrated in the double and triple pbp mutants ( data not shown ) . We observed an increase in the half-life ( >120 min ) for PBP2b28 compared with the wild-type protein ( 21 min ) , not only in pbp2b28 ( Fig . 5A ) but also in the triple pbp mutant ( Fig . 5C ) . Interestingly , we also detected an increased half-life for PBP1a28 ( >120 min ) and PBP2x28 ( 77 min ) in the triple pbp mutant compared with half-life of PBP1a ( 31 min ) and PBP2x ( 29 min ) displayed in the wild-type strain ( Figs . 5B and 5C ) . Our hypothesis is that these stability changes in the PBP mutants could have been responsible for the fitness/morphological alterations in the pbp2b mutant and the compensatory effects in the triple pbp mutant , and this will be discussed later . We also investigated the putative cause of a morphological change in the pbp2b28 mutant , and particularly in the cell division process . It was previously proposed that the life cycle of bacterial cells consists of repeated controlled enlargement , septum formation , and cell division [19] . In this scenario , FtsZ is an essential protein , which was postulated as the force generator that drives the cell division process , since the correct localization of all proteins involved in this mechanism is dependent on FtsZ [19] , [20] . To evaluate the impact of pbp mutations on cell division , we determined the FtsZ localization by immunofluorescence microscopy , using a polyclonal antibody against FtsZ . Before septum formation , the FtsZ division ring has been described to be localized at the mid cell , as we also observed for the wild-type strain ( Fig . 6A ) . However , an atypical FtsZ placement was found in pbp2b cells , with an apparent helical structure rather than the mid-cell localization found in wild-type cells . In the pbp2b28 pbp2x28 mutants , although we observed a coccoid morphology , the FtsZ localization was still altered . In contrast , a total FtsZ placement restoration was found in the triple pbp mutants ( Fig . 6A ) . The higher stability of PBP2b28 led us to speculate that this increased protein level could also cause a delocalization of this protein mutant . Therefore , we constructed the PBP2b-GFP and PBP2b28-GFP fusions , expressed ectopically from a multicopy plasmid , in order to study their localization in S . pneumoniae cells . We observed that PBP2b-GFP was localized equatorially in the wild-type strain as described previously [16] . However , PBP2b28-GFP revealed an atypical helical distribution , similar to FtsZ in the pbp2b28 mutant ( Fig . 6B ) , suggesting that PBP2b could interact with FtsZ . To investigate this , we analyzed this putative interaction by using a bacterial two-hybrid system ( Bacteriomatch II , Stratagene ) , confirming this hypothesis by the detection of a positive interaction between PBP2b28 ( or PBP2b ) and FtsZ , but not between PBP2b28 ( or PBP2b ) and PBP2x . This served as a control of specificity in addition to the positive and negative controls included in this system ( Fig . 7 ) .
Mutations associated with antibiotic resistance are known to impose a fitness cost in several bacterial species . Nevertheless , this physiological cost can be ameliorated by compensatory mutations in the same altered gene or in others involved in the developed resistance mechanism , thereby maintaining the same resistance level [21] . This compensatory phenomenon can occur with or without selective pressure . However , under antibiotic exposure , only those mutations able to maintain the resistance level will be selected , thus permitting the survival and persistence of resistant strains [22] . In S . pneumoniae , individual pbp2b or pbp2x mutations are considered primary βL resistance determinants that confer low-level resistance , whereas pbp1a mutations are acquired later and are responsible for increasing this level . In this work , only pbp2b mutations were related to a significant fitness cost , indicating that these mutations were a disadvantage when competing with wild-type strains . This finding indicates that the pbp2b mutants were only able to survive under antibiotic pressure due to their resistance . However , in the absence of antibiotics , the only way for these mutants to compete with wild-type strains is by improving their fitness by acquisition of compensatory mutations . In this work , one of the most important findings is that these compensatory mutations are also responsible for an increase in βL resistance . It is well known that the combination of pbp mutations that confers βL resistance is very complex , and that pbp alleles have DNA mosaics with mutations in the region that encodes the transpeptidase domain . These DNA mosaics may differ by over 20% from DNA sequences from sensitive pbp genes and from each other , resulting in changes in amino acid of more than 10% [23] . In this scenario with PBPs showing multiple mutations , it is very difficult to distinguish those involved in resistance development from the alterations caused by the natural evolution of the genes . Therefore , the contribution of each mutation is unclear , with the exception of a few conserved mutations in the transpeptidase domains of PBPs . In an attempt to identify the amino acids involved in the compensatory mechanism described here , different fragments of pbp1a28 and pbp2x28 were transformed into the pbp2b28 mutant , and the resistance and fitness cost were evaluated . Although two fragments from the pbp1a28 gene were able to confer resistance , only A2 compensated the fitness cost caused by the pbp2b28 mutation . The amino acid sequence deduced from the DNA sequence of A2 revealed that the E285Q substitution might be involved in the compensatory mechanism . On the other hand , when pbp2x alleles from different clinical strains were used to transform pbp2b28 , only one of them ( pbp2x52 ) was unable to ameliorate the fitness cost of this mutant . Four pbp2x alleles shared 9 substitutions compared with Cp1015 , with the exception of pbp2x52 , which differed in only 1 residue ( D278 in the place of N278 ) . This result suggests that N278 , present in the pbp2x mutations that compensated the fitness cost of pbp2b28 , could be also a compensatory mutation , at least in the set of mutations described for pbp2x12 , pbp2x52 , and pbp2x62 . Rieux and coworkers [11] studied the relationship between the acquisition of penicillin resistance and virulence , and demonstrated that the single and double pbp2b pbp2x mutants lost virulence in an intraperitoneal infection model in Swiss mice . After several passages in mice , the pbp2x mutations showed stability . However , for the pbp2b mutants , virulent revertants were recovered without intragenic modifications , suggesting that extragenic compensatory mutations were involved in this phenomenon . On the other hand , Trzcinski and coworkers [12] reported a rise in the βL resistance levels of the susceptible D39 strain , by the sequential transformation of the pbp genes obtained from clinical isolates . This event was accompanied by an increase in the fitness cost of a triple pbp mutant as determined by a rat nasal colonization model , but without growth alterations occurring in vitro . However , in the Trzcinski's work , no fitness compensation was detected among the pbp mutations obtained from isolates belonging to the Spain6B-2 , Hungary19A-6 , or serotype-9V Spain9V-3 international clones . In contrast with these reports , we showed that pbp2x and pbp1a genes carrying specific mutations that conferred βL resistance , not only compensated the pbp2b-associated fitness cost in Cp1015 and D39 strains , but also increased the βL resistance levels . We also demonstrated that this that this compensatory mechanism was present in the clinical strains that belonged to the new serotype-14 variant of the Spain9V-3 pneumococcal clone identified in Argentina [13] . Related to this , several other studies have reported that the fitness cost associated to antibiotic resistant mutants depends on the specific mutations related to the resistant mechanism and the genetic background [5] , [24] , [25] , [26] , [27] , [28] . Here , we also used strain D39 , the same strain used by Trzcinski and coworkers [12] , as an acceptor for pbp transformations . However , contrasting results were obtained , which might be explained by considering the different nature of the pbp mutations and the particular evolution of the clinical isolates carrying these mutations . These assumptions were supported by an amino acid sequence analysis of the PBPs encoded by the mutated pbp2b , pbp1a and pbp2x genes from the clinical strains used in this study , with and a clear difference being found with those described by Trzcinski and coworkers [12] . When the βL resistance of the double and triple pbp mutants was analyzed , we found an increase not only in its level , but also in its spectrum . We postulate a direct correlation between an increase in βL resistance and fitness cost compensation . A similar mechanism was previously shown for fluoroquinolone resistance , by analyzing the parC/E and gyrA mutations obtained from clinical pneumococcal strains using competitive fitness assays in vitro [8] . Recently , this phenomenon was also described for fluoroquinolone resistance in E . coli by fitness assays in vitro and in vivo using a laboratory strain [28] . Although the main focus of the present work was the compensatory evolution associated to the pbp mutations that conferred βL resistance , it is important to highlight another topic linked to PBPs , the cell division process . In particular , we studied the putative mechanism that caused morphological changes in pbp2b mutants . It is already known that PBPs play an important role in cell wall synthesis [29] , but morphological alterations have not been previously described for pbp mutants in the absence of β-lactams . In this work , we described an atypical transition from a coccoid to a rod-like shape and also other alterations caused by pbp2b mutated genes obtained from βLR clinical isolates , with this morphological phenotype being reproduced in different laboratory strains . In a previous work , it was proposed that PBP2b , an essential class-B PBP , is involved in peripheral peptidoglycan synthesis in S . pneumoniae [16] . Related to this , bacterial cell morphogenesis is regulated by a controlled peptidoglycan synthesis , where PBP activity is essential for the normal progression of this process [30] , and also by an orchestrated interaction of cytoskeleton proteins triggered by FtsZ polymerization [31] . The septal accumulation observed by Van-FL staining in the pbp2b mutants indicated an abnormal cell division . To investigate the impact of pbp2b mutations on this process , we analyzed the FtsZ localization by immunofluorescence in the pbp2b mutant and found an unusual localization pattern , revealing structures that resembled the helical Van-FL-stained sidewall shown by Bacillus subtilis [17] . To search for the reason why the pbp2b mutants showed these morphological alterations , we also measured the stability of PBP2b28 and found that PBP2b28 showed an increased half-life , in contrast with that observed in the wild-type strain . The fact that PBP2b28 displayed a higher protein level led us to investigate a putative delocalization of this protein in a pbp2b28 genetic background . Immunofluorescence assays showed that PBP2b-GFP fusion localized equatorially as FtsZ did in the wild-type strain described by Morlot and coworkers [16] . In contrast , PBP2b28-GFP fusion revealed an atypical helical distribution similar to FtsZ in the pbp2b28 mutant . It is possible that protein-protein interactions may have altered the localization of certain proteins , and we suspected that PBP2b28 could have interacted with FtsZ , thus modifying its normal emplacement . Using a bacterial two-hybrid system , we demonstrated that FtsZ was able to interact with PBP2b28 as well as with the wild-type protein , but not with PBP2x . Considering that PBP2b28 was delocalized , we propose that this interaction of PBP2b28 with FtsZ led to a misplacement of this protein , and consequently , to an aberrant cellular morphology that caused a decreased fitness . Moreover , these results support the idea that this interaction is essential for the control of cell morphogenesis in S . pneumoniae , as suggested for E . coli [32] , [33] . In addition to this phenomenon , we suspect that PBP2b28 presented a decreased transpeptidase activity , as all the PBP mutants confer βLR by a decreased affinity to these compounds [34] , [35] . Because PBP2b has been involved in the elongation stage during cell division [16] , we propose that as PBP2b28 could not elongate efficiently then these cells accumulated septa , as demonstrated by Van-FL staining assays , and this could be another cause of the morphological alterations found in the pbp2b28 mutant . In our laboratory , work is now in progress in an attempt to gain a better understanding of this complex process and this not well understood molecular mechanism . To try to explain the putative cause of compensation of the PBP2b28 alterations by PBP1a28 and PBP2x28 , we also analyzed the stability of both mutant proteins . We observed that PBP1a28 and PBP2x28 displayed an increased half-life in the triple pbp mutant as we had also for PBP2b28 . Considering that all these PBPs have transpeptidase activity , these results suggest that PBP1a28 and PBP2x28 have a complementary function on PBP2b28 , based mainly on their increased protein levels . We propose that this event could have caused the compensatory effect on fitness and the morphological alterations in the pbp2b28 mutant . In the current work , we reported compensatory extragenic mutations that restored the fitness alterations imposed by mutations related to βL resistance . Given that the fitness compensation determined by assays in vivo and in vitro were similar , our data suggest that the selection of fitter strains may have taken place in the natural habitat of S . pneumoniae . This phenomenon has an important clinical relevance , since these compensated pbp mutants cannot only maintain but also increase the βL resistance levels , in contrast with previous reports [22] . Our results also indicate that clinical strains acquired those pbp mutations which improved their resistance-associated fitness cost , producing a clear competitive and selective advantage , and thus raising the potential spreading of βL resistance . We think it is probable that S . pneumoniae exploited its high transformability to acquire the compensatory pbp mutations that are essential for its persistence and dissemination . This idea is supported by the experimental data , which indicated that the pbp mutants showed no transformability alterations , and that the pbp mutants could be selected at a transformation rate higher than 10-2/µg of DNA , thus demonstrating that compensatory mutations can be acquired in a one-step transformation . If this compensatory evolution was due to the sequential incorporation of adaptive mutations , then the occurrence of this phenomenon may increase rapidly under antibiotic pressure . Therefore , it is possible that certain mutations were selected for their compensatory effect on fitness in addition to their contribution in developing higher βL resistance levels . It is known that the emergence and stability of antibiotic resistance is a complex biological process , being driven by different factors such as the volume of antibiotic used , the rate of resistant mutant formation , the fitness cost imposed and the compensatory mechanisms to improve that cost [22] . Many studies on the effect of a reduction in β-lactam consumption have reported a sustained resistance level to S . pneumoniae [22] , suggesting that other factors in addition to antibiotic exposition are contributing to the persistence and the evolution of βL resistance . In the present work , the compensatory mechanism seems to be an important factor , favoring environmental long-term persistence and the spreading of βLR strains . Moreover , it is known that fitness compensation improves the dissemination ability of resistant strains , an essential trait that characterizes pneumococcal clones . However , we could not show that this phenomenon was particularly associated to the pbp genes for other successful international clones . As mentioned above , fitness restoration has not been reported for pbp mutations obtained from clinical strains that belong to the Spain6B-2 , Hungary19A-6 , and serotype-9V Spain9V-3 international clones [12] . Nevertheless , we demonstrated that a compensatory mechanism was present , at least in the new serotype-14 variant of the Spain9V-3 clone recently characterized in our laboratory [13] , and also in a serotype-9V reference strain of this clone ( ATCC 700671 ) isolated in France [36] . Therefore , this model contributes to the understanding of βL resistance evolution in S . pneumoniae .
The bacterial strains used in this work are listed in Table S1 in Supporting Information S2 . βLR pneumococcal strains were obtained from invasive infections of pediatric patients and belonged to the Spain9V-3 international clone [13] . Cells were routinely grown at 37°C in Todd-Hewitt broth supplemented with 1% bovine serum albumin . For antimicrobial susceptibility testing , strains were grown at 37°C in a 5% CO2 atmosphere on Mueller-Hinton agar with 5% defibrinated sheep blood . Penicillin , cefotaxime , and piperacillin MICs were determined by agar dilution following a CLSI protocol [37] . To amplify the transpeptidase region of the pbp1a , pbp2b , and pbp2x genes by PCR , we used the F1a/R1a , F2b/R2b , and F2x/R2x [13] primer pairs , respectively ( Fig . S1 in Supporting Information S1 ) . The internal fragments of the pbp1a gene were amplified and sequenced with the following primer pairs: Fa1/Ra1 ( from position 94 to 234 ) ; Fa2/Ra2 ( from position 599 to 1002 ) , Fa3/Ra3 ( from position 881 to 1294 ) , Fa4/Ra4 ( from position 385 to 521 ) and Fa5-Ra5 ( from position 476 to 629 ) ( Fig . S6 in Supporting Information S1 ) . The internal fragments of the pbp2x gene were amplified and sequenced with the following pair primers: Fx1/Rx1 ( from position 125 to 271 ) , Fx2/Rx2 ( from position 238 to 370 ) ; Fx3/Rx3 ( from position 320 to 469 ) and Fx4-Rx4 ( from position 445 to 738 ) ( Fig . S7 in Supporting Information S1 ) . The internal fragments of the pbp2b gene were amplified and sequenced with the following primer pairs: Fb1/Rb1 ( from upstream of pbp2b to 114 ) ; Fb2/Rb2 ( from position 69 to 256 ) ; Fb3/Rb3 ( from position 217 to 439 ) ; Fb4/Rb4 ( from position 374 to 557 ) and Fb5/Rb5 ( from position 502 to downstream pbp2b ) ( Fig . S17 in Supporting Information S1 ) . The primers sequences are detailed in Table S4 in Supporting Information S2 . PCR products were amplified using the following parameters: initial denaturation at 94°C for 4 min , 30 cycles of denaturation at 94°C for 45 s , annealing at 55°C for 30 s , elongation at 72°C for 1 min , and a final extension at 72°C for 10 min . Chromosomal DNA was isolated using the Wizard Genomic DNA purification kit ( Promega ) , following the manufacturer's instructions . The PCR products were amplified from chromosomal DNA and used to genetically transform the Cp1015 and D39 strains as described previously [38] . Transformants were selected on Mueller-Hinton agar plates supplemented with 5% defibrinated sheep blood containing 0 . 05 µg/ml piperacillin or 0 . 1–0 . 3 µg/ml cefotaxime . Natural competence was performed as described previously ( 3 ) . Strain Cp1015 and its respective pbp mutants were individually grown to the mid-logarithmic phase in Todd-Hewitt broth at 37°C . Cells were then washed three times and resuspended in PBS . In order to analyze cell shape modifications by flow cytometry analysis , bacteria were injected into a FACS Aria . Between 50 , 000 and 100 , 000 events were counted and analyzed by using WinMDI software . To compare populations , the data for each strain were plotted on a two-dimensional graph ( x-axis , forward scatter; y-axis , side scatter ) . Relative fitness was quantified as the average number of surviving progeny of a particular genotype , and this was compared with average number of surviving progeny of competing genotypes after a single generation . The wild-type strain genotype was normalized at wt = 1 and the fitnesses of other genotypes were measured with respect to that genotype [39] . The cost of a resistance mutation was determined by direct competition against the susceptible Cp1015 strain as described previously [7] . Individual strains were exponentially grown to an OD600 nm of 0 . 2 and cultures were diluted 2 , 000-fold . Mixed cultures containing equivalent amounts of the Cp1015 and pbp mutant cells ( about 5×104 CFU/ml ) were incubated in antibiotic-free medium for 6 h . These mixed cultures were then diluted 1 , 000-fold to avoid the typical lysis of S . pneumoniae cultures at the stationary phase , and cells were cultured for an additional 6 h . The number of viable cells was determined at 0 , 6 and 12 h by plating serial dilutions of the culture on BHI agar with 5% defibrinated sheep blood , containing 0 . 05 µg/ml piperacillin ( for pbp2b simple mutants ) , 0 . 1 µg/ml cefotaxime ( for pbp double and triple mutants ) , or with no antibiotic . The number of susceptible cells was calculated by subtracting the number of resistant cells from the total cell number revealed by the CFU counts of the plates without drug . To determine the CFU numbers , the mean of four counts was calculated . The number of generations of the resistant and Cp1015 strains in the mixed culture was calculated by using the following formula: ( log B - log A ) / ( log 2 ) , where A is the number of CFU/ml at time zero and B is the number of CFU/ml at the end of each cycle ( 6 h and 12 h ) . The relative fitness of each strain was determined from the ratio of the number of generations of the resistant strain and Cp1015 . The mean of four to nine replicate competition assays were determined . Statistical tests were performed using Instat software . These assays using a mice model were performed as described previously [40] , but with specific modifications for S . pneumoniae . Seven male C57BL/6 , 4 to 5 weeks old ( obtained from Comision Nacional de Energia Atomica , Ezeiza , Argentina ) , were inoculated intraperitoneally ( under isoflurane anesthesia ) with a mix of equal parts of D39 strain and an isogenic pbp mutant ( 1×105 CFU ) in 0 . 1 ml 50 mM glucose ( in PBS ) . These mice were sacrificed by CO2 asphyxiation after 2 days , and strains were recovered from each homogenized liver by plating onto BHI agar containing piperacillin , cefotaxime , or no antibiotic , as described above . Relative fitness was determined from the ratio of the number of generations of the resistant strains and strain D39 . ANOVA statistical tests were performed with Instat software . This study was carried out in strict accordance with the recommendations in the Guide to the Care and Use of Experimental Animals published by the Canadian Council on Animal Care . The experimental protocols performed were reviewed and approved by the Ethic Committee of the Facultad de Ciencias Químicas , Universidad Nacional de Cordoba ( Permit number 15-07-68964 ) . The Ethic Committee is constituted by the following professors of the Facultad de Ciencias Químicas-UNC: Laura Chiapello , Claudia Sotomayor , Margarita Briñon , Teresa Scimonelli , Santiago Quiroga , Mariana Contin and Graciela Granero . The protocols and the Animal Laboratory of our Institute ( CIBICI-CONICET , Argentina ) obtained an Animal Welfare Assurance from NIH , USA ( Assurance number A5802-01 , see http://grants . nih . gov/grants/olaw/assurance/500index . htm ? Country=AR#GridTop ) . All the clinical strains used in this work were provided by the strain collection of the Centro de Estudios Avanzados en Pediatría ( Córdoba , Argentina ) . The Ethics Committee declared in writing that no formal ethical approval was needed to use these clinically obtained materials , because the specimens were remnant from patient samples of blood and nasopharyngeal swabs collected from routine analysis in bacteriology laboratories of public hospitals , and the data were analyzed anonymously . Pneumococcal strains were routinely grown to an OD600 nm of 0 . 2 in Todd-Hewitt broth . Samples were collected and incubated with 2 µg/ml Van-FL ( Molecular Probes ) for 20 minutes at 37°C . Cells were centrifuged , washed three times with PBS and fixed with 3% paraformaldehyde . After several washes with PBS , cells were spotted on glass slides , air dried , dipped in methanol at −20°C for 10 min , and allowed to dry at room temperature . The cells were then stained with DAPI ( Sigma ) at a final concentration of 0 . 2 µg/ml for 10 min before being observed , and images were acquired using a Nikon Eclipse TE-2000 epifluorescence microscope fitted with a Nikon Digital Sight DS-U1 camera . This was performed using ACT-U software , and images were processed with Adobe Photoshop CS version . Cp1015 and the isogenic mutants pbp2b28 , pbp2b28pbp2x28 , and pbp2b28pbp2x28pbp1a were exponentially grown at 37°C in Todd-Hewitt medium , and samples for electron microscopy were then collected , centrifuged , and fixed with 4% formaldehyde-2% formalin in 0 . 1 M cacodylate buffer for 1 hour at room temperature . An additional fixation with 1% osmium tetroxide in cacodylate buffer was carried out for 1 hour at room temperature . These fixed cells were dehydrated using an increasing concentration of acetone , and embedded in polymerized Araldite at 60°C for 48 hours . Thin sections were obtained using a JEOL JUM-7 microtome equipped with a glass or gem grade diamond knife , and microphotography was performed with a Zeiss LEO 906E microscope . Exponentially growing cells ( OD600 nm of 0 . 25 ) were fixed with 3% paraformaldehyde for 15 min at room temperature and incubated for 45 min on ice . The fixed bacteria were washed three times in PBS ( pH 7 . 4 ) and resuspended in GTE ( 50 mM glucose , 20 mM Tris-HCl , 10 mM EDTA ) , with a freshly prepared lysozyme solution in GTE being added to a final concentration of 2 µg/ml . 10 µl samples were immediately distributed onto poly-L-lysine microscope slides and air dried . The slides were dipped in −20°C methanol for 5 min , in −20°C acetone for 30 s , and then allowed to dry completely . After rehydration with PBS , the slides were blocked for 1 hour at 37°C with 2% bovine serum albumin ( Sigma ) in PBS ( BSA-PBS ) . Cells were then incubated with a 1∶100 dilution in BSA-PBS of rabbit polyclonal anti-FtsZ antibody [41] for 1 hour at 37°C . After washing 10 times with PBS , samples were incubated with a fluorescein-conjugated secondary antibody in PBS ( Alexa Fluor 488 , Molecular Probes ) for 30 min at 37°C in the dark . After removing the secondary antibody by washing samples several times with PBS , DAPI was added at a final concentration of 0 . 2 µg/ml and the samples were incubated for 10 min at room temperature . Slides were washed again , allowed to dry , and then mounted using Dako fluorescent mounting medium ( Invitrogen ) . The final slides were stored at −20°C for several days . Images were acquired using a Nikon Eclipse TE 2000 epifluorescence microscope , and processed with Adobe Photoshop CS version . The pbp2b gene was amplified from Cp1015 wt and Cba-28 clinical strains with primers F2bf and R2bf ( Table S4 in Supporting Information S2 ) . The gfpmut3 gene was amplified from the pCM18 [42] plasmid using the primer pair Fgfp3/Rgfp3 ( Table S4 in Supporting Information S2 ) in order to obtain a gfp copy under the control of the pneumococcal constitutive promoter , Pc [43] . The pbp genes were cloned into pGEMT-easy ( Promega ) and the gfpmut3 gene into pCRTOPO 2 . 1 ( Invitrogen ) , thereby generating the pGEM-pbp and pGFPTOPO plasmids , respectively ( Table S4 in Supporting Information S2 ) . The pbp genes were exscinded by XhoI digestion of pGEM-pbp and inserted into the SalI-restricted pGFP-TOPO plasmid . The SalI site was located in pGFP-TOPO upstream the transcription origin of gfpmut3 and downstream the Pc . The resultant plasmid pGFP-TOPO-pbp contained a pbp2b N-terminal fusion to the gfpmut3 gene under the control of Pc . This construction was then removed from pGFP-TOPO-pbp by EcoRI digestion and inserted into the same restriction site in the E . coli-S . pneumoniae shuttle vector plasmid pAT18 [44] . The resultant vector was named pAT18pbp-gfp , and was used to transform the wild-type strain ( Cp1015 ) and the Cp1015 pbp2b mutants . The transformant selection was made in BHI agar plates supplemented with 5% blood sheep and 2 . 5 µg/ml erythromycin . All constructions were confirmed by PCR and sequencing with the same primers used for the amplification of each gene . Insertion duplication mutagenesis was used to construct strains expressing HA-tagged PBPs . Primer pairs F2btag/R2bHA , F2xtag/R2xHA and F1atag/R1aHA ( Table S4 in Supporting Information S2 ) were designed to amplify approximately the last 300–400 bp of the pbp2b , pbp2x and pbp1a genes fused to the sequence of the HA tag . The PCR products were cloned into the XhoI and XbaI sites of plasmid pEVP3 [43] , and the final constructions ( named pEVP32bHA , pEVP32xHA , pEVP31aHA ) were used to transform the strains Cp1015 , Cp1015 pbp2b28 and Cp1015 pbp2b28 pbp2x28 pbp1a28 . Selection of the transformants was made in BHI 5% sheep blood agar supplemented with chloramphenicol ( 2 µg/ml ) . All constructions were confirmed by PCR and sequencing with the external primers for amplified regions , as mentioned above . Protein expression was inhibited in exponentially growing cultures in Todd Hewitt broth with 0 . 5% Yeast Extract ( total volume 400 ml ) at an OD600nm of 0 . 2 by the addition of kanamycin ( 500 µg/ml ) . One-hundred ml aliquots were withdrawn at 0 , 30 , 60 and 120 min after the addition of the antibiotic . Cells were immediately harvested , washed once with 1 ml of ice cold PBS and resuspended in 400 µl of ice cold MQ water plus Complete protease inhibitor cocktail ( Roche ) , before being subjected to five freeze-thaw cycles followed by sonication using a Sonics VibraCell VCX130 sonicator ( 20 cycles , 30 s ON 30 s OFF , 40% amplitude ) . To solubilize the membrane proteins , 100 µl of 5X RIPA buffer ( 250 mM Tris pH 7 . 4 , 750 mM NaCl , 5%NP-40 , 2 . 5% Sodium deoxycholate , 0 . 5% SDS ) were added and the samples were heated for 10 min at 95°C , and cell debris was removed by centrifugation for 15 min at 20 , 000 g . The total protein content was determined by using bicinchoninic acid assay [45] . After sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transfer to nitrocellulose membranes of 25 µg of the protein extract , the membranes were probed with a mouse anti-HA primary antibody ( 1∶2 , 000; Abacam ) and with a goat anti-mouse immunoglobulin G secondary antibody conjugated to horseradish peroxidase ( 1∶2 , 500; Invitrogen ) . Detection was performed with an enhanced chemiluminescence substrate ( SuperSignal West Pico Chemiluminescent Substrate; Pierce ) and Hyperfilm CL film ( GE ) using exposures of between 1 and 10 min . Western Blot bands were quantified using the Gel-Pro Analyzer v3 . 1 software and the half-life was calculated with Graph Pad Prism v5 . 3 software . The bacterial two-hybrid system BacterioMatch II ( Stratagene ) was used to screen for interactions of PBP2b with FtsZ protein and PBP2x . The pbp2b coding sequence was amplified by PCR with the primer pair F2bdh/R2bdh ( Table S4 in Supporting Information S2 ) from the Cp1015 and Cba-28 strains , digested with XhoI/BamHI and cloned into the XhoI/BamHI restricted bacterial two-hybrid vector pBT . In addition , the coding sequences of ftsZ and pbp2x were amplified by PCR with F2xdh/R2xdh , FftsZdh/RftsZdh ( Table S4 in Supporting Information S2 ) , respectively , digested with XhoI/BamHI or XhoI/BglII , in that order , and cloned into the pTRG vector cleaved by XhoI/BamHI . The E . coli XL1 Blue MR , the BacterioMatch II Validation Reporter strain ( Stratagene ) , was co-transformed with the two plasmids . Growth cultures of the clones on M9 minimal media agar plates containing 2 . 5 mM 3-amino-1 , 2 , 4-triazol ( 3-AT ) and 10 µg/ml of streptomycin were assessed according to the manufacturer's instructions . | For many years , pneumococcal infections have been usually treated with β-lactams . However , the rapid emergence of β-lactam resistance has complicated the antimicrobial treatment of these infections in the last two decades . The emergence and stability of antibiotic resistance is a complex biological process driven by different factors , such as the volume of antibiotic used . Furthermore , many studies on the effect of a reduction in β-lactam consumption have reported a sustained resistance level to S . pneumoniae , suggesting that other factors contribute to the persistence of β-lactam resistance . By horizontal gene transfer , S . pneumoniae is able to acquire genes from resistant strains or the commensal streptococci , which confer β-lactam resistance . Here , we show that when certain resistance genes are acquired individually , an important cost results in the bacterial fitness . However , some clinical strains which have acquired genes that increase β-lactam resistance can also compensate the fitness cost imposed by this resistance , thereby producing a selective advantage and raising the potential spreading of β-lactam resistance . We suggest that pbp1a and pbp2x mutant alleles are acquired for their compensatory effect on fitness in addition to their contribution in developing higher β-lactam resistance levels , and that this process may occur even in the absence of antibiotics . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"medicine",
"infectious",
"diseases",
"streptococci",
"medical",
"microbiology",
"biology",
"microbiology",
"bacterial",
"pathogens"
]
| 2011 | Compensatory Evolution of pbp Mutations Restores the Fitness Cost Imposed by β-Lactam Resistance in Streptococcus pneumoniae |
Nonsyndromic hearing impairment ( NSHI ) is a highly heterogeneous condition with more than eighty known causative genes . However , in the clinical setting , a large number of NSHI families have unexplained etiology , suggesting that there are many more genes to be identified . In this study we used SNP-based linkage analysis and follow up microsatellite markers to identify a novel locus ( DFNA66 ) on chromosome 6q15-21 ( LOD 5 . 1 ) in a large Danish family with dominantly inherited NSHI . By locus specific capture and next-generation sequencing , we identified a c . 574C>T heterozygous nonsense mutation ( p . R192* ) in CD164 . This gene encodes a 197 amino acid transmembrane sialomucin ( known as endolyn , MUC-24 or CD164 ) , which is widely expressed and involved in cell adhesion and migration . The mutation segregated with the phenotype and was absent in 1200 Danish control individuals and in databases with whole-genome and exome sequence data . The predicted effect of the mutation was a truncation of the last six C-terminal residues of the cytoplasmic tail of CD164 , including a highly conserved canonical sorting motif ( YXXФ ) . In whole blood from an affected individual , we found by RT-PCR both the wild-type and the mutated transcript suggesting that the mutant transcript escapes nonsense mediated decay . Functional studies in HEK cells demonstrated that the truncated protein was almost completely retained on the plasma cell membrane in contrast to the wild-type protein , which targeted primarily to the endo-lysosomal compartments , implicating failed endocytosis as a possible disease mechanism . In the mouse ear , we found CD164 expressed in the inner and outer hair cells of the organ of Corti , as well as in other locations in the cochlear duct . In conclusion , we have identified a new DFNA locus located on chromosome 6q15-21 and implicated CD164 as a novel gene for hearing impairment .
Nonsyndromic hearing impairment ( NSHI ) is the most frequent hereditary sensory defect in humans worldwide . The condition is clinically and genetically extremely heterogeneous , with more than 160 loci identified today . Autosomal dominant NSHI ( ADNSHI ) shows great variation in age of onset , rate of progression , severity and frequencies affected in contrast to autosomal recessive NSHI ( ARNSHI ) that is usually congenital/prelingual and non-progressive [1] . Currently , around 30 causative genes for ADNSHI have been identified . These genes are involved in a wide variety of molecular processes such as gene regulation , cytoskeleton dynamics , cell-cell junction formation , endocytosis and membrane transport [2] . Additional causative genes are expected to be discovered , since over 20 loci have been mapped without the corresponding genes being identified , and novel loci and/or genes are regularly being uncovered ( http://hereditaryhearingloss . org ) [1 , 3] . In the clinical field , identification of these hearing loss genes has greatly aided genetic counselling on hearing impairment . With the advances in next-generation sequencing technologies it is now possible to quickly screen most known genes implicated in NSHI simultaneously either by using customized capture arrays for targeted genes or exome sequencing [3 , 4] for the benefit of families , where the causative mutation can be identified . For these cases , diagnosis as well as important predictive information for the remaining family members can be offered [5] . However , with the extreme genetic heterogeneity in NSHI , a large proportion of the screened families still have an unexplained etiology . In this study , we identified a novel locus ( DFNA66 ) for dominant inherited NSHI on 6q15-21 in a large Danish family . By the use of a custom capture array and next-generation sequencing , we searched for the causative mutation in the region and identified a nonsense mutation in CD164 [OMIM 603356] . The gene encodes CD164 , a small transmembrane sialomucin protein involved in adhesion , migration and endocytosis and we provide data on the variant- , gene- , and functional level implicating the gene in hearing impairment .
A multi-generational family from Denmark with ADNSHI , affecting 17 individuals in five generations ( Fig 1A ) participated in the study . Audiograms and audiological data were collected from 13 individuals born between 1931 and 2003 . The hearing impairment is moderate to severe ( Fig 1B ) . Age of onset varied from newborn ( detected through neonatal screening ) , age 3–6 or early twenties . The audiograms showed variable patterns with either a flat audiogram affecting all frequencies , or , at least initially , a basin shape with the most severe affection on the mid-frequencies . In some cases the hearing impairment remained stable , in others it progressed somewhat affecting a broader spectrum of frequencies over the years . Representative audiograms can be found in S1 Fig . One family member ( IV-21 , Fig 1A ) with hearing impairment had experienced severe recurrent otitis media in childhood . From careful assessment of his audiograms ( S1 Fig ) we were not able to unequivocally determine if his hearing impairment was conductive or sensorineural . His phenotype was therefore set to unknown ( grey pedigree symbol , Fig 1A ) . Initial sequencing of seven known hearing loss genes ( WFS1 [OMIM 606201] , GRHL2 [OMIM 608576] , EYA4 [OMIM 603550] , ACTG1 [OMIM 102560] , GJB2 [OMIM 121011] , MYO6 ( exon 25 ) [OMIM 600970] , and SLC26A4 [OMIM 605646] ) failed to identify any mutations , prompting us to perform a genome-wide linkage analysis to identify the responsible locus for the hearing impairment in the family . Eleven individuals were then selected for single nucleotide polymorphism ( SNP ) genotyping using the Affymetrix 50K Xba240 array . The genotyped individuals are indicated with yellow squares in Fig 1A . After quality control and SNP pruning , 11 , 034 markers in approximate linkage equilibrium were included in a parametric linkage analysis using an autosomal dominant model with full penetrance and allele frequencies obtained from the CEU population . A single 25 Mb genome-wide significant linkage peak was identified on chromosome 6q15-q21 ( LOD score = 3 . 6 ) , with the critical haplotype flanked by markers rs9294390 ( 88 , 556 , 380 bp ) and rs6910441 ( 113 , 518 , 576 bp ) ( hg19 ) ( S1 Table ) . This region contains 101 annotated genes ( S2 Table ) . The locus is relatively close to EYA4 ( DFNA10 ) [6] , however a meiotic cross-over in three affected individuals excluded DFNA10 as the cause of the hearing loss in this family , consistent with the initial sequencing where no variations were found in EYA4 . To validate and possibly narrow down the locus , 26 family members were genotyped for seven microsatellite markers across the locus . A multipoint linkage analysis was carried out with allele frequencies determined from all genotyped founders and penetrance set to 1 . In the analysis , the affection status was set to “unknown” for three individual in total . These were IV-21 , because of the uncertainty about the origin of his hearing impairment ( see description of the family ) and individual V-19 and V-38 because of their young age ( 16 and 10 years respectively ) being below the upper observed age of onset of the hearing impairment in this family . The analysis including 23 individuals mapped the locus between D6S462 ( 90 , 928 , 511 bp ) and D6S433 ( first marker outside region ) , thus narrowing down the locus by approximately 2 Mb in the proximal end and increasing the LOD score to 5 . 1 ( Fig 1C ) . The genomic position of the locus is 90 , 928 , 511 to 113 , 518 , 576 bp ( hg19 ) . In an attempt to identify the causal mutation , nine candidate genes within the linked region were Sanger sequenced: SOBP [OMIM 613667] and FOXO3 [OMIM 602681] , known to cause deafness in mice [7 , 8] , and seven other genes ( GJA10 [OMIM 611924] , POU3F2 [OMIM 600494] , FAXC ( also known as C6orf168 ) [no OMIM] , LIN28B [OMIM 611044] , Hsa-mir-587 [no OMIM] , AMD1 [OMIM 180980] , and LAMA4 [OMIM 600133] ) , selected based on homology to known hearing loss genes or expression in the inner ear . Only common sequence variations ( MAF above 1% in ESP6500 ) , unlikely to cause hearing loss , were identified in these genes . We then applied a NimbleGen customized targeted capture array and next-generation sequencing ( NGS ) in order to sequence the entire locus in one affected individual ( IV-31 ) ( Fig 1A ) . Statistics for the bioinformatics analysis can be seen in S4 Table . After uploading the VCF file to Ingenuity Variant Analysis , an initial filtering based on mapping quality and chromosomal position identified 28 , 200 variations across the entire locus . After filtering out common variants ( MAF above 1% ) , 1609 variants remained . Of these , two were found in coding regions; a variant c . 574C>T [NM_006016 . 4] in CD164 and rs143143212 in MMS22L . Filtering the 1609 variants for functional effect ( S2 Fig ) , one variant passed through the filter i . e . c . 574C>T in CD164 . The variant is predicted to cause a truncation of CD164 by introducing a premature stop codon at amino acid position 192 ( p . R192* ) [UniProtKB NP_006007 . 2] ) and is not present in any available databases . Genotyping of all 26 family members with DNA available confirmed that the CD164 mutation was found in all individuals carrying the critical haplotype ( Fig 1A and 1D ) . Genotyping of 1200 unrelated Danish control individuals for the c . 574C>T nonsense mutation did not identify anyone carrying the c . 574C>T variant . By genotyping 2400 control chromosomes from the same background population as the family , the power is 80% to detect a variant with a minor allele frequency as low as 0 . 001 , suggesting that the mutation is unlikely to be a rare polymorphism in the Danish population . To ask if other nonsense or frameshift mutations in CD164 had been reported , we searched all relevant , available databases . In dbSNP138 , we found a nonsense mutation ( rs11542733 ) which was originally submitted to dbSNP120 by a large-scale sequencing effort of expressed sequence tags in 2001 [9] . The mutation was reported in individual NA06993 ( CEPH 1341 . 13 ) . We obtained genomic DNA from this individual ( Coriell Cell Repositories , New Jersey , USA ) and by Sanger sequencing we were not able to confirm the presence of this mutation ( S3 Fig ) , suggesting that the record is likely due to an artefact from early high throughput sequencing . In conclusion , the c . 574C>T mutation is to our knowledge the first CD164 nonsense mutation identified in humans . To estimate the frequency of CD164 mutations among patients with unknown cause of hearing impairment , we sequenced all coding exons and splice junctions of CD164 using DNA samples from 46 independent index cases . The cases were 15 unrelated probands from Denmark ( the index patient from 12 families and 3 sporadic cases ) selected based on their hearing impairment phenotype with basin shaped audiograms , 25 index patients from the Netherlands based on phenotype with postlingual onset ( 1st or 2nd decade ) , progression of the hearing impairment and cookie-bite or flat audiogram configuration , and 6 probands of Pakistani families with ARNSHI that displayed linkage to chromosome 6 . The recessive families were included as several hearing impairment genes ( e . g . TMC1 [OMIM 606706] , TECTA [OMIM 602574] , MYO7A [OMIM 276903] ) have been found to underlie both autosomal-dominant and recessive NSHI ( http://hereditaryhearingloss . org/ ) . However , no sequence variants likely to cause hearing impairment were found , suggesting that mutations in CD164 are not a common cause of NSHI . CD164 contains seven coding exons and expresses a protein referred to as CD164 , MUC-24 or endolyn [10] . Five splice variants of the gene have been reported , with isoforms 1–3 encoding a membrane bound form by the use of the full exon 6 , and isoforms 4 and 5 encoding a soluble form of the protein by alternative splicing of exon 6 or the alternative use of exon 7 ( Fig 2A ) . Isoform 1 ( ENST00000413644 ) and 4 ( ENST00000310786 ) account for the vast majority of expressed transcripts across different tissues , found by the Genotype-Tissue Expression project ( GTEx ) [11] . As the c . 574C>T mutation is located at the end of exon 6 , the mutation is predicted to affect only the membrane bound forms of CD164 ( isoforms 1–3 ) . Isoform 1 encodes a 197 amino acid long protein with a large extracellular region with two heavily glycosylated mucin-like domains , separated by a cysteine-rich domain , a transmembrane domain , and a short cytoplasmic region containing a canonical YXXФ sorting motif ( where X stands for any residue and Ф for a large hydrophobic residue ) ( YHTL ) ( Fig 2B ) . As previously mentioned , the c . 574C>T mutation causes a substitution of an arginine ( R192 ) for a stop codon ( p . R192* ) , thereby deleting the last six amino acids of the CD164 C-terminus ( RNYHTL ) , including the sorting motif . An amino acid sequence alignment of CD164 from different species shows a 100% conservation of these six C-terminal CD164 residues from human to roundworm ( Fig 2C ) , indicating a high selective pressure against amino acid changes in this sequence , consistent with its role in subcellular trafficking of proteins to the lysosomal compartment in cells [12] . To assess the functional effect of the truncating mutation on sorting and localization of CD164 , we first studied the subcellular localization of the C-terminal region ( CTR ) of wild-type and mutant CD164 fused to fluorescent marker proteins . We co-transfected human embryonic kidney ( HEK ) -293 cells with plasmids encoding two fusion proteins: ( i ) an mCherry fluorescent protein N-terminally fused to the transmembrane segment and the CTR of CD164 ( mCherry-CD164-WT-CTR ) and ( ii ) an eGFP fluorescent protein N-terminally fused to the transmembrane segment and the CTR of CD164 lacking the last 6 amino acids ( eGFP-CD164-R192*-CTR ) ( Fig 3A ) . This was done to detect and distinguish the subcellular localization of wild-type and truncated CD164 C-terminal regions simultaneously in the same experiment . Using confocal microscopy , images of live cells were captured two days after transfection . This demonstrated that in the steady-state , the truncated fusion protein ( green ) was found mostly at the plasma membrane , while the wild-type fusion protein ( red ) was predominantly located in intracellular vesicles , suggesting a grossly abnormal sorting of the truncated fusion protein ( Fig 3B ) . Identical findings were obtained when cells were transfected with plasmids encoding the opposite combination of fluorescent marker proteins ( colour swap ) ( Fig 3C ) . In both dye swap experiments a small amount of truncated fusion protein was detected in the cytosol . Passive internalization is the most likely explanation for this because the truncated fusion protein was present at very high levels in the plasma membrane . To investigate if wild-type and R192* CD164 with intact extracellular domain would exhibit a similar trafficking difference as the C-terminal region , HEK cells were stably transfected with constructs encoding human full-length wild-type CD164 and the truncated CD164 R192* , respectively . A qPCR assay , able to distinguish wild-type and mutant transcripts and quantifying total CD164 , were used to select two cell lines expressing wild-type and mutant CD164 , respectively , at comparable levels ( S3 Table and S4 Fig ) . The assay showed that endogenous CD164 expression in the mutant cell line accounting for around 20% of the total CD164 expression . Due to the high amount of CD164 ( >95% ) in the endo-lysosomal system under normal steady-state conditions , and in order to observe the timing of the endocytic trafficking of wild-type and mutant proteins , all CD164 present at the cell surface on living transfectants were saturated with anti-CD164 antibodies at 0°C , as cooling arrests internalisation ( T0 ) . At T0 , CD164 was present at the plasma membrane in both cell lines , as expected ( Fig 4A and 4B ) . The fate of CD164 was then followed after raising the temperature to 37°C to initiate internalization . After 10 minutes , most of the wild-type CD164 was internalized ( Fig 4C ) with no further change in localization after 30 min ( T30 ) ( Fig 4E ) , indicating that wild-type CD164 was rapidly ( within minutes ) cleared from the cell surface and that no recycling of CD164 took place within this timeframe . In contrast , only low levels of CD164 R192* were internalized after 10 and 30 minutes ( Fig 4D and 4F ) . Untransfected HEK cells did not produce a CD164 signal over background in these stainings . This experiment demonstrated that CD164 R192* was trapped at the plasma membrane . Because CD164 has been shown to form disulfide-linked homodimers [13 , 14] , we speculated whether CD164 R192* could heterodimerize with CD164 WT . To this end , we generated expression constructs in which FLAG , HA or myc epitope tags were inserted at various positions in a relatively poorly conserved region immediately following the signal peptide of CD164 . We first tested the expression of various tagged constructs compared to their untagged counter parts by transient transfection in HEK cells followed by immunoblotting analysis using antibody to human CD164 . Untagged CD164 migrated as several bands with predominant species around 80–100 kDa under reducing conditions ( Fig 5A ) . This is consistent with previous studies reporting migration of reduced CD164 as several bands ranging from 60–100 kDa depending on the cell line or tissue analysed . This migratory behavior is believed to be due to extensive and variable glycosylation of CD164 molecules [13–15] . We found that CD164 R192* expressed at similar or slightly higher levels and with identical molecular size as wild-type CD164 , indicating that the mutation did not impair protein stability or glycosylation state . No signal was detected in empty vector transfected cells , showing that the endogenous CD164 was expressed at a low level compared to the exogenous CD164 in these experiments . The various epitope tags affected somewhat the CD164 expression level and the FLAG tag also the size distribution , with enhancement of species around 65 and 140 kDa , probably via effects on the glycosylation pattern . We next co-transfected HEK cells with distinctly tagged CD164 and CD164 R192* ( or empty vector ) in various combinations as indicated . Two days post-transfection , cells were lysed and wild-type or mutant CD164 immunoprecipitated using the appropriate anti-tag antibody , followed by immunoblotting for co-precipitation of the other CD164 form . This analysis showed that HA4-CD164 R192* was able to co-immunoprecipitate FLAG4-CD164 ( Fig 5B left upper panel ) . Upon swapping of the tags , FLAG4-CD164 R192* was co-immunoprecipitated with HA2-CD164 ( Fig 5B right upper panel ) . Control immunoblots demonstrated appropriate co-expression of the two constructs ( Fig 5B middle and lower panel ) . Thus , in our experiments mutant CD164 was able to co-precipitate wild-type CD164 and vice versa demonstrating that mutant CD164 can form heterodimers with wild-type CD164 in HEK cells . Given their ability to form heterodimers , we next tested if the internalization-deficient CD164 R192* mutant could negatively affect internalization of wild-type CD164 . We co-transfected HEK cells with HA4-CD164 R192* and FLAG4-CD164 followed by double-staining of the cells with HA and FLAG antibodies at 0°C ( Fig 6 ) . Under these conditions of arrest of the endocytic machinery both wild-type and truncated CD164 was localized at the plasma membrane ( Fig 6A–6C ) . However , after shifting the cells to internalization permitting conditions ( 37°C ) most of the wild-type CD164 was internalized after 10 min with no further change at 30 min , whereas the majority of CD164 R192* maintained localisation on the plasma membrane ( Fig 6D–6I ) . Thus , while these results support the findings on the internalization of wild-type and lack thereof for mutant CD164 presented in Fig 4 , they do not support the idea that mutant CD164 R192* negatively affects internalization of wild-type CD164 . It should be mentioned that in a minority of cells , we observed slow or no internalization of both wild-type and truncated CD164 . Although we cannot completely rule out an effect of mutant CD164 , we believe this observation is more likely explained by a non-functional internalization system in these cells . Given the large effect of the p . R192* mutation on CD164 subcellular trafficking in our cell based assays , we speculated whether the transcript containing the mutation was expressed in cells from the affected family members . In mammalian cells , transcripts containing premature stop codons are generally degraded by nonsense-mediated mRNA decay ( NMD ) . The efficiency of NDM , however , depends on the exact position of the premature stop codon [16] . We extracted RNA from a blood sample from the index patient ( IV-5 , Fig 1A ) and after RT-PCR using intron spanning primers and Sanger sequencing , we aligned the obtained sequence to the human genome using BLAT to validate that it was from cDNA and not from genomic DNA ( Fig 7A ) . We found that both the normal and mutated CD164 transcripts were expressed in peripheral blood cells ( Fig 7B ) , demonstrating that the CD164 c . 574C>T transcript escapes NMD . This is consistent with the “55 bp rule” described for NMD , where the surveillance system in general seem to fail to distinguish premature stop codons if they are positioned in the last exon or in the second to last exon and located less than 55 bp from the final intron [17] , which is the case for the present mutation . For the gene to have a likely role in disease pathology , it should be expressed in the relevant tissue . From the publicly available BioGPS [18] database CD164 transcripts appear to be widely expressed across different tissues in the human body , with high expression levels in the thyroid , whole blood , colon and small intestine , and medium expression in many other organs and lowest expression levels in the brain [18] . CD164 transcripts are also expressed in the human fetal cochlea , according UniGene Hs . 520313 , with inner ear data derived from Morton Human Fetal cDNA Library [19] . The detailed cellular distribution of CD164 at the protein level within the inner ear has however not been determined [20] . The protein expression pattern of cd164 in the inner ear was therefore investigated by staining of sections of mouse cochlea at postnatal day five using two different antibodies ( Fig 8 and S5 Fig ) . This analysis indicated cd164 expression in the cochlear neurons , inner and outer hair cells of the organ of Corti , cells of Kolliker’s organ , cells in the lateral cochlear wall behind the spiral prominence and cells of the stria vascularis . The two antibodies showed the same expression pattern in the cochlea . The expression in the hair cells was weaker than in the other cell types , consistent with the mRNA expression pattern of cd164 in the Shared Harvard Inner-Ear Laboratory Database ( SHIELD ) database .
In this study , we mapped a novel locus ( DFNA66 ) for NSHI to chromosome 6q15-21 . The locus contained FOXO3 and SOBP , known to cause deafness in mice , but Sanger sequencing and careful assessment did not identify any variation in these genes . By targeted sequence capture combined with NGS we instead identified a novel nonsense mutation in CD164 , which was the only rare variant with a predicted functional impact , and thereby the best candidate in the region . In our filtering strategy we did not filter solely on the presence in dbSNP , because with the increasing number of pathogenic variants being submitted to public databases , this may lead to low frequency causal variants being missed . Instead , we chose a conservative minor allele frequency threshold of 1% , which is a rather conservative threshold when performing mapping studies of high penetrant rare variants in Mendelian disorders [21] . In the family , the nonsense mutation segregated in all affected individuals , as well as to a 10-year old girl reported to be unaffected from multiple audiological examinations during her early childhood ( Fig 1A , individual V-38 ) . Interestingly , in a recent audiological follow-up after the finding of the mutation , a small dip in mid frequencies in her audiogram was found , which could be the first signs of an effect of the mutation , consistent with the broad range of age of onset observed in the family for the trait . If her hearing impairment progresses , all individuals carrying the variant will then display the phenotype , suggesting a high to complete penetrance with variable age of onset . Individual IV-21 , who was not included in the linkage analysis due to recurrent severe otitis media in childhood , did not have the mutation . We therefore concluded that his hearing impairment is likely caused by the many ear infections . In the search for rare disease causing mutations with high impact , linkage is an effective method for eliminating large fractions of the genome , but segregation and rarity alone is not sufficient to implicate a specific variant as pathogenic . In this study we therefore performed a number of functional studies to characterise CD164 and the effect of the mutation . The YHTL motif , deleted by the c . 574C>T nonsense mutation , is a canonical sorting motif known to be recognized by specific adaptor proteins in the cytosol , leading to subcellular trafficking of the transmembrane protein to endosomes and lysosomes [22] . In many transmembrane receptors ( e . g . mannose 6-phosphate receptor and sortilin ) the sorting motif mediates direct transport between the trans-Golgi network and endosomes , due to interaction with AP1 [22] . For other transmembrane proteins like CD164 and CD1 cellular trafficking to lysosomes also depend on AP3 , but through different routes . Whereas newly synthesized CD1 seems to be captured by AP3 in the TGN for direct sorting to lysosomes , CD164's lower affinity for AP3 , combined with a sorting signal residing in the luminal/extracellular domain , results in direct transport to the cell surface [23] . At the plasma membrane , the YHTL motif is recognized by AP2 and CD164 is subsequently rapidly endocytosed into early endosomes , a process known as the indirect route [10] . This is consistent with our functional data showing that CD164 R192* lacking the sorting motif is accumulated on the cell surface . Such a dramatic effect on localisation of CD164 when perturbing the YHTL sorting motif has also been seen in other cell types , where point mutations of the critical tyrosine ( Y ) and leucine ( L ) residues in the YHTL motif were shown to lead to retention of CD164 at the plasma membrane [10 , 23 , 24] . We are the first to study the effect of an YHTL-disrupting CD164 mutation identified in humans . Taken together , the data suggest that abnormal trafficking of CD164 is consistently observed across different cell types when the YHTL sorting motif is perturbed . The molecular mechanism through which truncated CD164 causes hearing loss is currently unknown . We have shown that the c . 574C>T mutant transcript is not degraded by NMD in whole blood in patients , and that CD164 R192* is able to dimerize with wild-type CD164 . We have also shown that CD164 R192* is trapped at the plasma membrane , but that the truncated protein does not appear to hold back wild-type CD164 on the surface in HEK cells , arguing against a direct dominant negative effect on wild-type CD164 internalization . However , it is possible that CD164 R192* may suppress other functions of wild-type CD164 via dimerization to cause hearing loss . It is also possible that the increased amounts of CD164 R192* protein at the plasma membrane could exert a “toxic” effect in cells in the inner ear . Other organ-specific diseases arising as a consequence of alterations in the sorting signals of individual plasma membrane proteins has been reviewed in [25 , 26] . CD164 has been shown to regulate CXCR4 signaling in hematopoietic precursor cells [27] and myoblasts [24] . However , none of the affected family members was evaluated for hematopoietic disorders . Previously , CD164 have been studied in Drosophila and recently in zebrafish . In a study from 2006 , Zhou et al . found that endolyn-deficient Drosophila mutants were arrested in embryonic and early larval development [28] , and that a proportion of the growth-inhibited cells were undergoing apoptosis , suggesting a role for CD164 in cell proliferation . More recently , Mo et al . , studied the kidney function in zebrafish embryos after morpholino knockdown of endolyn expression , and found that despite the pronephric kidney appeared morphologically normal , clearance of fluorescent dextran injected into the common cardinal vein was delayed , suggesting a defect in the regulation of water balance in the morphant embryos [29] . Interestingly , the authors found that the defects could be rescued by expression of rat endolyn , but not by expression of endolyn lacking the canonical YXXФ sorting motif , suggesting that correct kidney function require endolyn endocytosis at least in zebrafish [29] . In the present family there were no reports of renal disease . The creatinine and carbamide levels , measured in peripheral plasma in one of the affected family members , were found to be within normal range , and no microscopic kidney abnormalities were reported in an autopsy report of an affected family member , deceased in 2008 . The phenotype of the morpholino zebrafish may however still be of some interest , since both the kidney and the inner ear contain polarised epithelial cells important for maintenance of fluid homeostasis . Furthermore , cd164 expression was detected in the stria vascularis ( among other key functional sites ) of the mouse cochlea , supporting the possibility of a role in endolymph homeostasis . As fluid homeostasis is important for correct hearing , this could be one possible mechanism through which CD164 is involved in hearing loss . In conclusion we have identified a novel locus for hearing impairment with LOD score 5 . 1 and identified CD164 as the most likely causative gene in the locus . Our data points towards an important role of CD164 in the function of the inner ear and suggest that the lack of the YXXФ motif , which is important for AP2 mediated endocytosis , underlies the hearing impairment in this family , however the exact molecular disease mechanism needs to be further investigated .
The proband was ascertained and the family pedigree constructed in collaboration between Department of Clinical Genetics , Vejle Hospital and Department of Audiology , Bispebjerg Hospital . One male ( IV-5 ) , with hearing impairment first diagnosed at about age 10 y , was examined several times . The audiograms at age 31 and at age 35 were similar , with 40 dB HL at 500 Hz , increasing to 70 dB HL at the frequencies 1000–4000 Hz , and improving to 20 dB HL at 8000 Hz ( Fig 1B ) . His daughter ( V-39 ) was diagnosed at neonatal hearing screening and carefully followed . She had at age 5 a sloping audiogram with 30–40 dB HL at frequencies 250–500 Hz , and 60–70 dB HL at 1000–2000 HZ and 50–60 dB HL at 4000–8000 Hz ( S1 Fig ) . Between age 5 and 6 , no progression was observed . A male in another branch of the family ( V-24 ) , experienced hearing impairment from the age of 3 , and at age 6 an audiogram showed a basin shaped curve with 30dB HL at 500 HZ , dipping to 60 dB HL at 1000 HZ and 40 dB HL at 4000 HZ . At age 19 , his audiogram showed 50 dB HL at 500 Hz , and a 60 dB HL at 1000–8000 HZ , thus illustrating progression ( S1 Fig ) . Vestibular complaints were not reported subjectively . Individual IV-21 had symptoms of hearing impairment and numerous purulent childhood middle ear infections > 20 punctures of the eardrum , culminating with an operation for choleastoma , which is a known complication of middle ear infection . From his audiogram ( S1 Fig ) it was not possible unequivocally to determine if his hearing impairment was sensorineural or conductive ( caused by the infections ) . His phenotype was considered unknown through the study . Genomic DNA was extracted from peripheral blood samples . Ten affected and one unaffected individual ( indicated in yellow in Fig 1A ) were genotyped using the Human Mapping 50K SNP Xba240 Array ( Affymetrix , High Wycombe , UK ) . Genotypes were called using the Genotyping Console ( Affymetrix ) and uploaded to the BCSNP data management platform ( BC Platforms , Espoo , Finland ) . Data on a total of 58 , 958 markers was generated . Those markers with Mendelian errors , which were detected with MERLIN , were removed from the dataset ( 491 markers ) . Removal of monomorphic markers and LD pruning ( using a sliding window of 50 SNPs and a r^2 threshold of 0 . 5 ) was performed using PLINK resulting in a filtered dataset of 11 , 034 markers in approximate linkage equilibrium with each other . MERLIN was also used to identify unlike genotypes , resulting in the removal of 221 genotypes from the dataset . Parametric linkage analysis was carried out with Merlin using an autosomal-dominant mode of inheritance with complete penetrance and a disease gene frequency of 0 . 0001 , SNP allele frequencies from CEU and genetic distances from the Affymetrix 100K Marshfield cM map . A follow-up analysis was performed by genotyping 26 available family members with seven microsatellite markers ( D6S1595 , D6S1644 , D6S1613 , D6S462 , D6S416 , D6S432 , and D6S433 ) positioned within and just outside the linked region from the SNP analysis ( S3 Table ) . Primer sequences were retrieved from the NCBI UniSTS database After PCR , the fragments were shipped to Eurofins Genomics ( Ebersberg , Germany ) for fragment analysis . Alleles were uploaded to BCSNP and parametric linkage analysis was performed with Mega2 [30] and SimWalk2 [31] , which can handle large pedigrees . Allele frequencies were calculated from founders . Due to the variable age of onset of the hearing impairment in this family , the affection status of two apparently healthy children ( 16 years and 10 years old , respectively ) was set to unknown . Similarly for one affected individual with multiple ear infections during childhood . Thus 23 individuals contributed to the follow-up linkage analysis . Disease allele frequency was set to 0 . 0001 and penetrance to 1 . All intron-exon boundaries and coding exons were sequenced for nine genes ( GJA10 , POU3F2 , C6orf168 , LIN28B , Hsa-mir-587 , SOBP , FOXO3 AMD1 , and LAMA4 ) . For POU3F2 , we were able to PCR amplify , but not to Sanger sequence through a highly GC rich region ( 98% GCs ) encoding a total of 21 glycine ( Gly ) residues in exon 1 . Attempts to sequence this GC rich region ( chr6:99 , 282 , 960–99 , 283 , 007 ) were performed by Sanger sequencing of two different PCR products , as well as providing the purified PCR product to Eurofins Genomics for direct Sanger sequencing using their custom service for difficult templates . As the same difficulty was found in two affected and two healthy control individuals , we assume that the failure is likely caused by polymerase failure and not by a mutation in the family . To exclude the presence of a trinucleotide expansion in this region , we amplified the region using a fluorescence-labeled primer pair followed by fragment length analysis at Eurofins Genomics . This analysis yielded a single peak for all samples analyzed ( four affected , four control individuals ) , excluding that the sequencing failure across this region was caused by a trinucleotide expansion . Oligo sequences are listed in S3 Table . A custom designed sequence capture array covering chr6:88 , 511 , 939–113 , 377 , 048 ( hg19 ) was obtained from NimleGen ( Roche NimbleGen , Madison , WI , USA ) . Genomic DNA from individual IV-31 ( Fig 1A ) was sheared by nebulization and universal adaptor oligonucleotides were ligated to the DNA . After this step , in order to enrich for the specific 6q region , the library was hybridized to the custom capture array . After washing to remove unhybridized material , captured molecules are recovered by heat-based elution and subjected to PCR amplification . The target-enriched library was quantified and subjected to deep sequencing on an Illumina Genome Analyzer , GAII using 36 bp reads . One lane of the flow cell was used for the sample . The raw sequence reads were aligned to the reference genome ( hg19 , NCBI build 37 ) using Burrows-Wheeler Aligner ( BWA ) [32] . This generated a total of 3 . 8 Gb of sequence . In order to identify single nucleotide variants and indels Genome Analysis Toolkit ( GATK ) was used described in “Best Practice Variant Detection with the GATK v4” [33] , which included removal of duplicate reads , local realignment around indels and base quality score recalibration before calling of genetic variants [34] . The sequencing depth and summary mapping statistics of the target region ( S4 Table ) were calculated using BEDTools [35] , PICARD ( http://picard . sourceforge . net ) , SamTools [36] and custom scripts . SNVs and indels were called using GATKs Unified genotyper [34] and subsequently SNVs were filtered in order to exclude SNVs with low mapping quality , low coverage and/or low quality scores . All variants passing this QC were indicated as PASS in the VCF file . The VCF file was uploaded to Ingenuity Variant Analysis for variant filtering . The filtering steps were ( 1 ) kept PASS upstream pipeline filtering AND kept that are on chromosome 6 AND between positions 88556380 and 113518576 , ( 2 ) excluded that are observed with an allele frequency greater than or equal to 1 . 0% of the genomes in the 1000 genomes project OR greater than or equal to 1 . 0% of the public Complete Genomics genomes OR greater than or equal to 1 . 0% of the NHLBI ESP exomes ( All ) ( 3 ) kept that are Frameshift , in-frame indel , or stop codon change OR Missense OR disrupt splice site upto 2 . 0 bases into intron OR structural variant ( S2 Fig ) . We used Ingenuity Variant Analysis version 3 . 0 . 20140520 Content versions: Ingenuity Knowledge Base ( Arrakis 140408 . 002 ) , COSMIC ( v68 ) , dbSNP ( Build 138 ( 08/09/2013 ) ) , 1000 Genome Frequency ( v3 ) , TargetScan ( v6 . 2 ) , EVS ( ESP6500 0 . 0 . 21 ) , JASPAR ( 10/12/2009 ) , PhyloP hg18 ( 11/2009 ) , PhyloP hg19 ( 01/2009 ) , Vista Enhancer hg18 ( 10/27/2007 ) , Vista Enhancer hg19 ( 12/26/2010 ) , CGI Genomes ( 11/2011 ) , SIFT ( 01/2013 ) , BSIFT ( 01/2013 ) , TCGA ( 09/05/2013 ) , PolyPhen-2 ( HumVar Training set 2011_12 ) , Clinvar ( 02/11/2014 ) . The CD164 c . 574C>T genotyping assays were developed by TIB MOLBIOL ( Berlin , Germany ) for the LightCycler 480 instrument ( Roche , Hvidovre , Denmark ) . Oligo sequences are listed in S3 Table . Genotyping was performed on 26 members of the Danish family and 1200 Danish control individuals ( 500 medical students from Aarhus University and 700 anonymous Danish blood donors ) . No information on the hearing ability of the control individuals was available . PCR primers were designed to amplify exons and surrounding intronic regions of the 7 exons of CD164 ( RefSeq nos . NM_006016 . 4 and NM_001142404 . 1 ) . Primer sequences are available in S3 Table . PCR conditions are available upon request . In total 46 individuals were screened for CD164 mutations . Among the tested individuals were the probands from five consanguineous Pakistani families with presumed recessive NSHL displaying linkage compatible with a locus on chromosome 6 . These five hearing impaired probands were from families DEM4010 ( LOD score 2 . 70 ) , DEM4026 ( LOD score 2 . 13 ) , DEM4028 ( LOD 1 . 23 ) , DEM4059 ( LOD score 3 . 00 ) and DEM4446B ( LOD score 2 . 54 ) . The human embryonic kidney cell line , HEK-293 ( cat . no . CRL-1573 , American Type Culture Collection , Boras , Sweden ) was maintained and cultivated according to standard techniques [37] . Transiently transfected cells were obtained by means of X-tremeGENE 9 ( Roche Applied Science , Hvidovre , Denmark ) transfection experiments following the manufacturer’s instructions using 1 . 5 μg total plasmid DNA and 9 μl X-tremeGENE 9 transfection reagent . In brief , HEK cells were seeded in 35 mm glass bottom microwell dishes ( MatTek , Ashland , MA , USA ) , and the next day they were co-transfected with pcDNA3 . 1-mCherry-CD164-WT-CTR and pcDNA3 . 1-eGFP-CD164-R192*-CTR . Stable transfected cells HEK cells were generated in T75 flasks using a total of 11 . 25 μg DNA ( pcDNA3 . 1-CD164-WT-Zeo , pcDNA3 . 1-CD164-WT-Hyg or pcDNA3 . 1-CD164- R192*-Neo ) and 33 . 75 μl X-tremeGENE 9 transfection reagent and selection of transfected cells were done using medium containing antibiotics ( Zeocine 100 μg/ml ( Invitrogen ) , Hygromycin 100 μg/ml ( Invitrogen ) , or Neomycine ( G418 ) 1 . 5 mg/ml ( VWR , Herlev , Denmark ) ) . Approximately one week after initiation of the selection procedure , non-transfected cells were dead and several positive clones were harvested after an additional week of the selection . Expression of CD164 was validated either by fluorescent microscopy of fluorescent marker genes ( mCherry and eGFP ) , immunostaining of CD164 or by qPCR . For live imaging of CD164 fusion proteins , HEK cells were co-transfected in glass bottom 35mm dishes ( MatTek ) with pcDNA3 . 1-CD164-WT-CTR-mCherry and pcDNA3 . 1-CD164R192*-CTR-eGFP . Two days post transfection the medium was replaced with DMEM without phenol red and live pictures was captured on a confocal laser scanning microscope ( LSM 780 , Zeiss , Jena , Germany ) using 63× water-immersion objective with a NA of 1 . 2 . Immunostaining and internalization was performed essentially as previously described [38] . In brief , stable transfected HEK cells or HEK cells co-transfected with FLAG4-CD164-WT and HA4-CD164-R192* seeded on glass were incubated on ice for 10 min to stop the endocytic machinery and subsequently incubated on ice for 90 min in medium containing 5 μg/ml purified mouse anti-human CD164 antibodies ( cat . no . 551296 , BD Biosciences ) , or a mixture of monoclonal anti-FLAG M2 antibodies ( cat . no . F3165 , Sigma ) and rabbit anti-HA antibodies ( cat . no . H6908 , Sigma ) . One fraction of the cells ( designated T0 ) were fixed in 4% paraformaldehyde ( Lillies buffer ) ( Buch & Holm , Herlev , Denmark ) for 15 min at RT , and permeabilized with PBS containing 0 . 25% ( w/v ) Saponin ( Sigma-Aldrich ) . The remaining cells were incubated further at 37°C in complete medium ( without antibody ) for 10 and 30 min , respectively . At the indicated time points cells were washed , fixed , and permeabilized as described above . Detection of CD164 in the stable transfected HEK cells was performed using secondary Alexa Fluor 488 goat anti-mouse antibody ( 1:400 , cat . no . A11029 , Invitrogen , Taastrup , Denmark ) . Detection of FLAG- and HA-tagged CD164 was obtained by using secondary Alexa Fluor 488 goat anti-mouse antibody ( 1:400 , cat . no . A11029 , Invitrogen ) and Alexa Fluor 568 donkey anti-rabbit antibody ( 1:400 , cat . no . A10042 , Invitrogen ) , respectively . Nuclei were stained with 4´ , 6-Diamidino-2-phenylindole ( Sigma-Aldrich ) and mounted on SuperFrost glass slides ( Hounisen , Risskov , Denmark ) . Sequential imaging was done on a confocal laser scanning microscope ( LSM 780 , Zeiss , Jena , Germany ) using 40× oil-immersion objective with a NA of 1 . 3 . HEK-293 cells in 35 mm plastic dishes were transiently transfected with untagged or epitope-tagged CD164 and CD164 R192* or empty pcDNA3 . 1 vector using X-tremeGENE 9 , as described above , and cultured for 2 days . For CD164 protein expression analysis , cells were thereafter lysed in reducing SDS-PAGE sample buffer and subjected to immunoblotting using sheep anti-human CD164 primary antibody ( AF5790 ) and horseradish peroxidase-coupled anti-sheep secondary antibody ( HAF016 ) , both from R&D Systems . For CD164 dimer formation analysis , cells were solubilized in immunoprecipitation buffer , as described [39] . Cell lysates were then incubated with 2 μg antibody to the HA tag ( 12CA5 clone ) and immune complexes were precipitated using protein G agarose beads ( 16–266 , Millipore ) . Aliquots of the immunoprecipitates or the pre-immunoprecipitation lysates were subjected to SDS-PAGE under reducing conditions followed by immunoblotting with horseradish peroxidase-coupled antibodies to the FLAG tag ( Sigma-Aldrich A8592 , M2 clone ) or the HA tag . Secondary antibodies were detected by chemiluminescence ( SuperSignal West Femto , #34095 , Pierce ) . A qPCR assay to detect the ratio between wild-type and mutant transcripts as well as total expression of CD164 in the double transfected cell lines was developed . Primers were designed to amplify total CD164 transcripts ( recognising both transcripts ) as well as the mutated and wild-type transcript ( allele specific primers ) . For each cell lines RNA was extracted from cell pellets using RNeasy ( Qiagen ) and cDNA was synthesized using iScript cDNA Synthesis kit ( BIO-RAD ) and 500 ng input RNA . Minus RT reactions were included to control for genomic DNA contamination . qPCR with was carried out for the transfected cell lines as well as untransfected HEK cells for control . The geometric mean of three genes ( ACTB , HTRP and TBP ) was used to normalize for cDNA content . All reactions were performed in triplicates . Fold changes were calculated relative to untransfected HEK cell . The relative amount of mutated and wild-type transcript within each cell line was calculated by taking the ratio of each transcript level to the level of total CD164 transcripts . Total RNA from peripheral blood lymphocytes was isolated from one of the affected family members ( Fig 1A , IV-5 ) using the PAXgene Blood RNA System consisting of a blood collection tube ( PAXgene Blood RNA Tube ) and nucleic acid purification kit ( PAXgene Blood RNA Kit ) ( Qiagen ) . The RNA was reverse-transcribed onto cDNA by using HT11V primers and the Superscript II kit ( Invitrogen ) . RT-PCR was carried out with forward and revers primers positioned in exon 5 and 6 respectively , thereby spanning intron 5 ( NM_006016 . 4 ) ( S3 Table ) . The PCR product was sequenced on both strands using Sanger sequencing and aligned to the CD164 gene using the BLAT program ( BLAST like alignment tool ) . Three wild-type mice at postnatal day five from the albino C57BL/6J-Tyrc-Brd inbred strain were used for the expression analysis . The heads of all samples were dissected in PBS before fixation for two days in 10% formalin at 4°C , washing , dehydrating and embedding in paraffin wax . Embedded samples were cut into 8μm thick sections along the sagittal plane . Immunohistochemistry was then carried out according to the manufacturer’s instructions on slides using the Ventana Discovery machine with the manufacturer’s reagents CC1 ( cat . no 950–124 ) , EZPrep ( cat . no 950–100 ) , LCS ( cat . no 650–010 ) , RiboWash ( cat . no 760–105 ) , Reaction Buffer ( cat . no 95–300 ) , and RiboCC ( cat . no 760–107 ) . The DABMap Kit ( Ventana; cat . no 760–124 ) with hematoxylin counterstain ( cat . no 760–2021 ) and bluing reagent ( cat . no 760–2037 ) were used . All antibodies were diluted in ‘Antibody staining solution’: 10% fetal calf serum , 0 . 1% Triton , 2% BSA and 0 . 5% sodium azide in PBS . The primary antibodies used were anti-CD164 ( SantaCruz , sc-33124 , 1:75 and St . John’s Laboratory , STJ92095 , 1:500 ) . The secondary antibody used was Jackson ImmunoResearch biotin-conjugated donkey anti-rabbit ( 711-065-152 , 1:100 ) . The stained slides were examined and images obtained using an AxioCam HRc camera mounted on a Zeiss microscope . The Hereditary Hearing loss Homepage ( http://hereditaryhearingloss . org ) OMIM—Online Mendelian Inheritance in Man ( www . omim . org/ ) PICARD ( http://picard . sourceforge . net ) SMART database ( smart . embl-heidelberg . de ) NetOGlyc 4 . 0 Server ( http://www . cbs . dtu . dk/services/NetOGlyc/ ) dbSNP ( http://www . ncbi . nlm . nih . gov/SNP/ ) 1000 Genomes project ( http://www . 1000genomes . org ) Exome Variant Server database ( http://evs . gs . washington . edu/EVS/ ) UCSC Genome Browser ( http://genome . ucsc . edu ) BioGPS ( http://biogps . org ) GTEx ( http://www . gtexportal . org/home/ ) Morton Human Fetal Cochlea cDNA Library EST Data ( http://brighamandwomens . org/Research/labs/BWH_Hearing/Cochlear_ESTs . aspx ) SHIELD: Shared Harvard Inner-Ear Laboratory Database ( https://shield . hms . harvard . edu ) | It is known that hearing impairment running in families can be caused by mutations in more than eighty different genes . However , there are still families where the responsible gene is unknown . By studying a large Danish family with dominant inherited hearing impairment , we found that the disorder cosegregates with genetic markers on chromosome 6 , suggesting that the responsible mutation lies within this chromosomal region . By sequencing this genetic locus , we discovered a mutation in the CD164 gene that is passed on to all the affected individuals . In the mouse ear , we demonstrated that the CD164 protein is expressed in hair cells and other sites known to be important for correct hearing . The identified mutation is predicted to result in shortening of the protein , leading to loss of an evolutionary conserved sequence important for cellular trafficking of CD164 . Using cell lines , we show that the truncated protein is trapped on the cell surface while the normal protein is internalized . This finding is important because it implicates for the first time a role for CD164 in the complex physiological processes of hearing and suggests that failed endocytosis may be a possible disease mechanism for some types of hearing impairment . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| []
| 2015 | A Novel Locus Harbouring a Functional CD164 Nonsense Mutation Identified in a Large Danish Family with Nonsyndromic Hearing Impairment |
A significant increase in microcephaly incidence was reported in Northeast Brazil at the end of 2015 , which has since been attributed to an epidemic of Zika virus ( ZIKV ) infections earlier that year . Further incidence of congenital Zika syndrome ( CZS ) was expected following waves of ZIKV infection throughout Latin America; however , only modest increases in microcephaly and CZS incidence have since been observed . The quantitative relationship between ZIKV infection , gestational age and congenital outcome remains poorly understood . We characterised the gestational-age-varying risk of microcephaly given ZIKV infection using publicly available incidence data from multiple locations in Brazil and Colombia . We found that the relative timings and shapes of ZIKV infection and microcephaly incidence curves suggested different gestational risk profiles for different locations , varying in both the duration and magnitude of gestational risk . Data from Northeast Brazil suggested a narrow window of risk during the first trimester , whereas data from Colombia suggested persistent risk throughout pregnancy . We then used the model to estimate which combination of behavioural and reporting changes would have been sufficient to explain the absence of a second microcephaly incidence wave in Bahia , Brazil; a population for which we had two years of data . We found that a 18 . 9-fold increase in ZIKV infection reporting rate was consistent with observed patterns . Our study illustrates how surveillance data may be used in principle to answer key questions in the absence of directed epidemiological studies . However , in this case , we suggest that currently available surveillance data are insufficient to accurately estimate the gestational-age-varying risk of microcephaly from ZIKV infection . The methods used here may be of use in future outbreaks and may help to inform improved surveillance and interpretation in countries yet to experience an outbreak of ZIKV infection .
A substantial body of experimental and clinical evidence implicates Zika virus ( ZIKV ) infection in the sharp rise in the incidence of microcephaly cases in Brazil at the end of 2015 . [1–5] Previous population-level studies investigating the relationship between ZIKV and microcephaly incidence found consistent patterns of high first-trimester risk and lower risk later in pregnancy , which is consistent with early clinical findings for ZIKV-associated microcephaly . [6 , 7] However , clinical studies investigating the link between ZIKV infection and a distinctive pattern of congenital abnormalities , collectively termed congenital Zika syndrome ( CZS ) , suggest that adverse outcomes are associated with ZIKV infection throughout pregnancy . [8–10] How these clinical findings link to the complex picture portrayed by the epidemiological data in Brazil is still unclear and leaves a substantial knowledge gap for those counseling pregnant women in ZIKV-affected populations . For example , why did the majority of Latin America demonstrate a relatively small rise in microcephaly incidence rates compared to those seen in Northeast Brazil , and why was the second wave of microcephaly in Brazil much smaller than the first despite two similar waves of Guillain-Barré Syndrome ( GBS ) ? [11] It is useful to consider the conceptual model in which observed population-level CZS incidence reflects underlying ZIKV transmission dynamics and a gestational-age-varying risk of CZS given infection with ZIKV . If a pregnant woman is infected at some point during gestation , her baby may present as a case of CZS with a probability conditional on the gestational age of her baby when infection occurred . Characterizing this link between underlying gestational risk of CZS and its presentation in epidemiological data has two potential benefits . First , an estimate of the underlying gestational risk profile from surveillance data may provide evidence to inform women of childbearing age to help plan pregnancy and mitigate exposure risk . [12] Second , if the risk profiles differ substantially between populations , those differences could support the study of alternative hypotheses of risk factors for CZS beyond ZIKV infection . For example , prior infection with another arbovirus has been suggested as a potential cofactor for risk of GBS , however prior arbovirus infection has not yet been shown to play a role in increased neurological adverse event risk . [13] Research on other teratogenic pathogens shows the potential importance of gestational age to CZS risk . [14] For example , prospective cohort studies of pregnant women have shown that infection early in gestation greatly increases the risk of congenital rubella syndrome and cytomegalovirus-associated adverse fetal outcomes relative to infection later in pregnancy . [15 , 16] Although a similar pattern seems likely to be the case for CZS , the timing and magnitude of risk throughout pregnancy remains uncertain . However , quantifying this underlying gestational-age-varying risk profile should be possible given reliable data on infection and CZS incidence combined and a robust statistical approach . Here , we demonstrate how a transmission model fitted to reported incidence data can be used to infer the relationship between gestational age at the time of ZIKV infection and the risk of microcephaly . We adapt this model to explore potential explanations , including changes in reporting rates and abortions , for the lack of a second observed wave of microcephaly incidence in Brazil .
We searched the literature , Pan American Health Organization ( PAHO ) , World Health Organization ( WHO ) and Brazilian state health authority websites for reports of suspected or confirmed ZIKV infection incidence and microcephaly cases in 2015 and early 2016 , building on a comprehensive literature search performed in 2016 . [17] In particular , we searched www . paho . org , www . who . int , Brazilian state-level ministry of health websites ( eg . www . suvisa . ba . gov . br ) , and PubMed for the terms “zika” and “microcephaly” . Where confirmation status of cases was not recorded ( eg . where incidence was only shown as “reported cases” ) , we classified these data as “notified” cases . Where suspected and confirmed cases were distinguished , we classified the sum of suspected and confirmed cases as “notified” cases . A summary of data included in the analyses can be found in S1 Table . Coverage , case definitions and protocols for ZIKV and microcephaly surveillance in Brazil and Colombia changed throughout 2015 and 2016 , making direct comparison of incidence data between these years difficult . [18] Although the first outbreak of laboratory confirmed ZIKV was reported in Brazil on 07/05/2015 , ZIKV reporting only became compulsory through the national notifiable information system ( SINAN ) on 17/02/2016 . [19 , 20] Furthermore , laboratory confirmation was only performed on suspected cases from previously unaffected areas and on certain subpopulations of interest ( eg . pregnant women , hospitalised patients with neurological complications ) . Reporting of microcephaly and other congenital abnormalities was through the information on live births system ( SINASC ) until November 2015 , when the new public health events registry ( RESP ) was implemented to improve surveillance in pregnant women and newborns . [21] In Colombia , passive reporting of ZIKV cases and major congenital abnormalities ( including microcephaly ) is ongoing through the National Health Institute ( INS ) surveillance system . [22 , 23] Laboratory testing and mandatory reporting of ZIKV infections based on clinical symptoms began on 14/10/2015 . [24] However , RT-PCR confirmation was not systematic and , like Brazil , was only used to confirm the presence of ZIKV in municipalities that had not yet detected the virus and in subpopulations at particular risk of complications . Therefore , the incidence of confirmed ZIKV infections gives an indication of the spread of the virus , but not necessarily the true magnitude and dynamics of the epidemic within affected locations . Some data sources were only available in graphical form , and these numbers were therefore extracted using a web digitiser ( https://automeris . io/WebPlotDigitizer/ ) . The results presented in the main text used data from: Northeast Brazil; Colombia; the city of Salvador , Bahia , Brazil; and state-reported incidence from Bahia , Rio Grande do Norte and Pernambuco , Brazil . Numbers of live births were obtained for Brazil from the SINASC/CGIAE/SVS/MS system . [7 , 25] For Colombia , live births were obtained from a publication of microcephaly and ZIKV infection incidence in Colombia and from Colombian ministry of health vital statistics . [23 , 26] We developed a two-component model to describe the relationship between the incidence of ZIKV infection and the incidence of microcephaly-affected births , depicted in Fig 1 . Full details of the model and sensitivity analyses are available in S1 Text and all code and analyses are available as an R package ( https://github . com/jameshay218/zikaInfer ) . Our aim was to estimate the shape and size of the risk window for developing ZIKV-associated microcephaly given infection during gestation , and to test for differences in inferred risk using data sets from various Brazilian states and Colombia . The first component of the model described the transmission dynamics of ZIKV via the Aedes aegypti mosquito vector based on the Ross-MacDonald model for vector-borne disease . [27] Through estimation of the force of infection over time , we estimated a per capita risk of human infection per unit time , PI ( t ) . The second component of the model described the risk of a fetus developing microcephaly given that the mother was infected in a particular week during pregnancy using a modified gamma function , P m ′ ( t ) . The expected proportion of microcephaly-affected births at any time t ( Fig 1E ) was obtained by multiplying these two components together: P m ( t ) = ∑ i = t - 40 t P I ( i ) P m ′ ( i - t + 40 ) ( 1 ) where Pm ( t ) is the probability of a ZIKV-associated microcephaly birth at time t , PI ( i ) is the probability of an individual becoming infected at time i ( and not before ) , and P m ′ ( i - t + 40 ) is the probability of a fetus developing microcephaly given ZIKV infection at gestational week i − t + 40 . Including a baseline microcephaly rate gives the probability of observing any microcephaly case at time t as: P m i c r o ( t ) = ϕ i [ P m ( t ) + P b - P b P m ( t ) ] ( 2 ) where Pb is the baseline per birth microcephaly incidence rate and ϕi is a multiplicative factor for the number of true cases that were reported in location i ( less than one indicates underreporting , greater than one indicates overreporting ) . Model parameters governing ZIKV transmission were obtained from the literature as described in S2 Table . Fixed parameters were mainly chosen based on a previously published transmission model resulting in a generation time ( ie . the length of time between the time of infection in a human case and the times of infection in the secondary human cases resulting from that case ) of approximately 20 days . [17] Given a fixed generation time , the model allowed the shape of the incidence curve for number of infected individuals ( ie . the number of individuals entering the I compartment of the SEIR model ) to vary depending on the value of the basic reproduction number , R0 . As R0 is comprised of multiple correlated parameters , all components of R0 other than the vector density per human were fixed . Vital statistics ( human life expectancy and population size ) for particular locations are described in S3 Table . We calculated the combined likelihood of observing ZIKV and microcephaly incidence data conditional on the model parameters , as described in S1 Text . Where full ZIKV infection incidence data were not available ( ie . Pernambuco , which only reported 3 weeks of ZIKV infection incidence in the first half of 2015 ) , we fit the two-component model to microcephaly incidence data alone and placed a 4-month-wide window on the time of peak ZIKV infection incidence given by the model as a uniform prior centered on the time of peak reported ZIKV incidence in Pernambuco ( ie . the peak of the three reported weeks ) . We fit the model separately to available notified and confirmed incidence data for each location using a Markov chain Monte Carlo ( MCMC ) framework written in R and C++ with ordinary differential equations solved used the rlsoda package . [28 , 29] By examining the posterior distributions of the parameters that determined the gamma risk profile , we also estimated the following parameters of interest: first week of gestational age where ZIKV infection confers a risk of microcephaly greater than 1 in 1000; last week of gestational age where ZIKV infection confers a risk of microcephaly greater than 1 in 1000; number of gestational weeks spent at risk; mean per-trimester risk; gestational week of greatest risk ( gamma mode ) . We carried out a sensitivity analysis using seroprevalence data from the city of Salvador in Bahia , Brazil to infer the microcephaly risk profile without the SEIR component of the model . Here , we assumed that reported acute exanthematous illness ( AEI ) was proportional to the true incidence of ZIKV infection during this time , and scaled the weekly reported incidence to give a final attack rate of between 59 . 4% and 66 . 8% in line with ZIKV IgG seroprevalence estimates for Salvador in May 2016 based on an NS1 antigen ELISA . [30 , 31] We included four model parameters to quantify potential changes in behaviour and reporting rates that could explain the two seasons of observed data in Bahia , Brazil , where only one wave of microcephaly incidence was observed despite two waves of ZIKV infection incidence . Time-dependent changes in behaviour and reporting were likely during the epidemic due to media hype and public awareness , demonstrated by changes in Google search behaviour shown in Fig 2A . First , we assumed that microcephaly reporting became 100% accurate from March 2016 ( the most recent change in case definition in Brazil ) and estimated the relative reporting rate prior to this as a model parameter . [18 , 32] Second , we assumed that immediately following the WHO declaration of a Public Health Emergency of International Concern in February 2016 , the rate of aborted pregnancies under 24 weeks gestation could have changed . [33–35] Third , we assumed that the number of ZIKV-affected births after this date may have changed , either due to avoided pregnancies or additional precautions taken by pregnant women to avoid infection relative to the rest of the population . [34 , 35] Finally , we assumed that ZIKV infection reporting accuracy may have changed after 11/11/2015 , when the Brazilian Ministry of Health declared a National Public Health Emergency , and just before WHO/PAHO issued an alert with laboratory detection guidelines for ZIKV . [36 , 37] Our model did not explicitly include seasonality , and the SEIR model was therefore only suitable for the single-season analyses . We therefore did not use the SEIR model component in the multi-season analysis , but rather assumed that the per capita risk of becoming infected with ZIKV was proportional to reported ZIKV infection incidence and that reported incidence of ZIKV infection represented a fraction of true cases scaled by one parameter up to 11/11/2015 and another from 11/11/2015 onwards .
Although our study data were from similar reporting systems , Fig 3 illustrates substantial differences in the key features of both ZIKV infection and microcephaly incidence patterns . Peak timing , width of the incidence curves , maximum per capita incidence and the lag between ZIKV infection and microcephaly incidence peaks differed by location and dataset . Variation in total and maximum per-birth microcephaly incidence indicates location-specific differences in the proportion of pregnant women that were infected with ZIKV or in the probability of developing microcephaly following infection . For example , weekly notified microcephaly incidence peaked at 32 . 6 cases per 10 , 000 births in Colombia , which was far lower than peak notified microcephaly incidence in Pernambuco , Brazil at 760 cases per 10 , 000 births , suggesting that microcephaly risk given infection and/or ZIKV attack rates were higher in Pernambuco . The lag between incidence peaks also varied , ranging from 23 weeks ( bootstrapped confidence intervals: 19-32 weeks ) for Colombia compared to 31 weeks ( bootstrapped confidence intervals: 30-36 ) from state-level reports for Bahia , Brazil . Given that only a small fraction of the Colombian population is at risk of arbovirus infection compared to the Brazilian population due to differences in vector ecology , it is unsurprising that the absolute per capita incidence of ZIKV infection and microcephaly are lower in Colombia than in Brazil . [22] However , as the population at risk of ZIKV infection is the same population at risk of microcephaly , we would expect the lag and relative magnitudes of ZIKV infection and microcephaly incidence to be the same between Brazil and Colombia . These differences in time lags and relative magnitudes therefore suggest that the time of peak risk during pregnancy may have varied between locations , potentially through differences in additional risk factors in these locations such as prior arbovirus exposure . Differences in observed incidence patterns may also arise as a result of reporting bias , which may be reduced using confirmed rather than notified case data . Although some confirmed microcephaly case data were available for Brazil , the only available data of confirmed microcephaly cases in Colombia reported number of cumulative confirmed cases . [39] It should also be noted that only a subset of suspected cases were laboratory confirmed in Colombia to test for the presence of ZIKV in municipalities which had yet to confirm ZIKV infection , and case reporting was otherwise based on clinical symptoms . [24] Due to variation in confirmation delays , we were unable to extract the time of birth of these cases and were therefore unable to use these data in model fitting . Between 03/01/2016 ( epidemiological week ( EW ) 1 of 2016 ) and 05/06/2017 ( EW 18 of 2017 ) , approximately 70% of notified microcephaly cases were discarded in Colombia ( 328 cases confirmed , 874 discarded , 37 under investigation ) , highlighting that total notified case data likely overestimates true incidence . [40] The proportion of total notified cases that were discarded was similar for Rio Grande do Norte ( 138 confirmed vs . 475 notified ) and somewhat higher for Pernambuco ( 365 confirmed vs . 2117 notified ) . Confirmed ZIKV infection incidence were available for Colombia but not Brazil due to the lack of reporting infrastructure during the first wave . [41] The lag between peak ZIKV infection and microcephaly incidence did not change when using notified or confirmed ZIKV infection or microcephaly data . We inferred interpretable gestational risk profiles for 5 of the 6 datasets used here ( Fig 4 ) . Though clear estimates of the gestational-age-varying risk were obtained for each location , substantial differences are apparent in the inferred gestational age of peak risk , duration of the risk period , and maximum absolute risk . The model did not produce a biologically interpretable risk profile using data from Pernambuco , Brazil that was comparable to those inferred using the other 5 datasets . Sensitivity analyses excluding the ZIKV infection incidence data from the model fitting for other locations were able to produce plausible risk profiles , suggesting biases in reported microcephaly incidence data for Pernambuco that could not be explained by the model ( see Section 4 . 3 , S1 Text ) . It is important to note that microcephaly is only one manifestation of CZS , and the risk profile of other adverse outcomes may differ . These risk estimates therefore apply only to the specific outcomes of proportionate and disproportionate microcephaly , which were not distinguished in these data . [8] The time from the peak of ZIKV infection incidence to the peak of microcephaly incidence indicates the typical gestational age at which microcephaly cases were infected . When using both state and city level notified ZIKV infection and microcephaly incidence from Bahia , Rio Grande do Norte , and Salvador , Brazil , the peak week of gestational-age-varying risk was estimated to be in the middle of the first trimester ( Fig 4 ) . When notified ZIKV infection incidence in pregnant women and confirmed microcephaly incidence from Northeast Brazil were used , the estimated peak gestational-age-varying risk was towards the end of the first trimester . Notified case data from Colombia were also suggestive of peak risk in the first trimester . Inference did not change substantially using confirmed as opposed to notified ZIKV infection incidence data for Colombia; however , using confirmed microcephaly incidence data for Rio Grande do Norte resulted in a shift of the risk profile towards the start of the second trimester . When a ZIKV infection incidence peak in a 4-month window around March 2015 was assumed for Pernambuco , the inferred microcephaly risk profile was highly skewed towards the first week of pregnancy , suggesting that these data are incompatible with the other 5 data sets . There was substantial variation in the inferred window of heightened gestational risk between different populations . The window of heightened gestational risk is estimated from the relative durations of the ZIKV infection and microcephaly incidence curves ( using an illustrative threshold of 1 in 1 , 000 infections leading to microcephaly to define the heightened gestational risk window ) . A narrow period of ZIKV infection incidence preceding a wide period of microcephaly incidence suggests a wide window of heightened gestational risk . If the period of heightened gestational risk is long , then infections at a particular point in time would present as cases of CZS across a wider interval of birth dates . Inferred risk profiles using notified case data from Rio Grande do Norte , the city of Salvador and Colombia all suggested heightened risk throughout pregnancy ( Fig 4 ) . Conversely , two similarly narrow ( or wide ) periods of ZIKV infection and microcephaly incidence would suggest a relatively small window of heightened gestational risk , as all ZIKV-affected pregnancies would present as births after a similar delay . A true microcephaly incidence period that is narrower than the ZIKV infection incidence period should not be possible , as the narrowest microcephaly incidence curve would arise when all infected pregnant women give birth after the same delay . Aggregated confirmed case data from Northeast Brazil , state-level notified case data from Bahia and state-level confirmed case data from Rio Grande do Norte all suggested a more limited window of risk during pregnancy , with lower risk suggested towards the end of pregnancy ( Fig 4 , Northeast Brazil , Bahia and Rio Grande do Norte ( confirmed cases ) ) . Public awareness , media hype , changing criteria for case reporting and variation in laboratory testing capacity likely resulted in changing reporting rates throughout the epidemic . [18 , 41 , 48 , 49] Location-specific time-varying changes in reporting sensitivity and specificity are therefore one potential explanation for differences in the risk profiles inferred using data from Northeast Brazil and Colombia . Given that Colombia was expecting an increase in microcephaly cases during 2016 , an increase in notified cases may have been reported before a true increase in confirmed cases , which would falsely suggest some gestational risk late in pregnancy . Time-varying reporting bias may also explain the extremely narrow and early window of risk inferred using data from Pernambuco ( Fig 4 , Pernambuco , Brazil ) . The impact of reporting bias is clearly demonstrated by the contrasting results using confirmed or notified microcephaly case data for Rio Grande do Norte , wherein confirmed data suggested a narrower and later risk window than the notified data . The absolute risk of CZS is more difficult to estimate as it depends on the true incidence of ZIKV infection in pregnant women and CZS cases as a proportion of live births . A high ZIKV infection attack rate with known microcephaly incidence would suggest a lower microcephaly risk per infection to the fetus than a low infection attack rate with the same observed microcephaly incidence . [7] Reported infection incidence data may be subject to under-reporting and over-reporting , potentially through missing asymptomatic or mild cases that might not present to surveillance systems ( under-reporting ) , or misclassifying infections caused by other arboviruses as ZIKV infection , namely dengue and chikungunya virus ( CHIKV ) ( over-reporting ) . [41 , 50] These confounders present identifiability problems in inferring levels of true incidence and therefore microcephaly risk; surveillance data in a scenario of high risk with under-reporting would be similar to a scenario of low risk with over-reporting . For example , during the 2015 wave in Brazil many cases of illness likely caused by ZIKV were misclassified as dengue infection , resulting in under-reporting of ZIKV infection incidence . [41] Over-reporting of microcephaly incidence during the initial wave of cases was also possible , due to changing case definitions , reclassification of suspected cases and increased awareness in surveillance systems . [18 , 32] Estimating the proportion of true ZIKV infections that led to observed microcephaly cases is therefore dependent on knowing the true risk of ZIKV infection during the epidemic period . ZIKV IgG seroprevalence was estimated to have reached 63 . 3% ( 95% confidence interval , 59 . 4 to 66 . 8% ) in Salvador , Brazil between 2015 and 2016 despite only 16 , 986 reported cases of AEI from a population of nearly 3 million ( approximately 0 . 6% ) , suggesting that under-reporting of ZIKV infection incidence was a key problem in this location . [30 , 31] By assuming that 100% of true microcephaly cases were reported but that reported ZIKV cases represented only a fraction of the true incidence , we inferred the absolute risk of ZIKV-associated microcephaly from each of the datasets ( S4 Table ) . The average first trimester risk of microcephaly given ZIKV infection was estimated to be 2 . 81% ( mean; 95% credible interval ( CI ) : 2 . 51-3 . 16% ) based on data from Bahia , Brazil , but much lower in the second trimester at 0 . 365% ( mean; 95% CI: 0 . 0715-0 . 588% ) . Conversely , the level of absolute risk estimated using notified case data from Colombia suggested that the risk was lower but consistent throughout gestation at 0 . 303% ( mean; 95% CI: 0 . 239-0 . 367% ) , 0 . 268% ( mean; 95% CI: 0 . 228-0 . 322% ) and 0 . 186% ( mean; 95% CI: 0 . 135-0 . 232% ) in the first , second and third trimesters respectively . The former estimate is slightly higher than risk estimates inferred based on seroprevalence data from French Polynesia which suggested a risk of 0 . 95% ( 95% confidence interval; 0 . 34–1 . 91% ) in the first trimester , whereas the latter estimate suggests a lower risk . [6] We performed a sensitivity analysis with better constraint on the true ZIKV attack rate by taking microcephaly and AEI data from Salvador , Brazil for 2015 scaled by recent ZIKV IgG seroprevalence data , as described in Section 6 , S1 Text . [30] Here , we assumed that the true risk of ZIKV infection in Salvador was proportional to the per capita reported incidence of AEI scaled such that the overall attack rate was between 59 . 4% and 66 . 8% . [31] Based on the ZIKV infection and microcephaly incidence data from Salvador , Brazil , we estimated the mean first trimester risk of microcephaly given ZIKV infection to be 3 . 06% ( mean , 95% CI: 2 . 66-3 . 49% ) ; the mean second trimester risk to be 0 . 805% ( mean , 95% CI: 0 . 649-0 . 980% ) ; and the mean third trimester risk to be 0 . 0833% ( mean , 95% CI: 0 . 0407-0 . 142% ) . We did not scale incidence data for any other location due to the lack of seroprevalence data . However , given that the model is powered by the pattern of microcephaly incidence relative to the pattern of ZIKV infection incidence after accounting for differences in infection risk and reporting , these risk estimates may apply to other locations if no additional cofactors affect the risk of microcephaly given infection . Despite a clear second wave of GBS incidence at the beginning of 2016 , no second wave of microcephaly incidence in Northeast Brazil was observed in the latter half of 2016 . [11] Similar to [11] , Fig 2B illustrates the incidence of microcephaly that would have been expected in Bahia , Brazil using our model framework and based on reported ZIKV infection incidence under the assumption that the underlying gestational-age-varying risk profile and reporting behaviour did not change from 2015 to 2016 . We used the population-level data fitting framework described above to test the hypothesis that plausible changes in behaviour or reporting are sufficient to provide a consistent narrative between the two waves of ZIKV and microcephaly case data . Fig 2A describes the timings of particular events that may have led to these changes . We considered four hypotheses describing changes in behaviour and reporting rates . First , we assumed that microcephaly reporting accuracy may have been different before week 11 of 2016 ( 13/03/2016 , the most recent change in case definition in for microcephaly reported through the Registro de Eventos em Saúde Pública ( RESP ) database in Brazil ) [32 , 51] and estimated the relative reporting rate for microcephaly prior to this that would be consistent with the observed data . Second , we assumed that immediately following the National Public Health Emergency announcement by the Brazilian Ministry of Health on 11/11/2015 , the frequency of early abortions ( up to 24 weeks gestation ) due to early detection of CZS may have increased . [36] The earliest date at which targeted abortions would be observed as a drop in birth rate would be 16 weeks after this shift in behaviour ( 02/03/2016 ) . [34 , 35 , 52] A reduction in birth rate from delayed pregnancy would also be possible; however , this would only appear approximately 40 weeks after the behavioural shift . Third , we assumed that the number of pregnant women affected by ZIKV after this date may have changed through additional precautions taken to avoid infection relative to the rest of the population . [53] Finally , we assumed that ZIKV reporting itself may have changed on 11/11/2015 before the start of the second wave of ZIKV infection incidence through increased surveillance , increased awareness and/or increased misclassification of other arbovirus infections as ZIKV infection . Over both time periods , we assumed that the per capita risk of becoming infected with ZIKV was proportional to reported ZIKV infection incidence , but that the scale of that proportion changed on 11/11/2015 following the potential change in ZIKV infection reporting . Based on state-level reports from Bahia , Brazil and assuming that ZIKV infection reporting did not change , our analyses suggest that the lack of a second microcephaly peak could be explained by the combined effect of: a 151% reporting rate of microcephaly cases prior to 13/03/2016 relative to fixed 100% accurate reporting after 13/03/2016; targeted abortions ending 88 . 4% of microcephaly-affected pregnancies prior to 24 weeks gestation; and a relative decrease in infection probability in pregnant women of 0 . 60% ( values shown are the maximum a posteriori probability ( MAP ) estimates ) . It is important to note that many of these parameters are highly correlated , suggesting that these data could be explained by a combination of multiple mechanisms , or by a greater contribution of some mechanisms and a reduced effect from the others ( Fig 5 ) . If ZIKV infection reporting accuracy increased substantially between the two waves in addition to the behavioural changes described above , then a smaller increase in the proportion of terminated pregnancies would have been necessary . Similarly , targeted abortions and precautions to avoid infection by pregnant women would present a similar reduction in microcephaly incidence , and these estimates are therefore highly correlated ( Fig 5C ) . Assuming that there were no targeted abortions , no additional precautions to avoid infection taken by pregnant women , and no change in microcephaly reporting accuracy , we estimated that these data could be explained solely by a 18 . 9-fold ( mean , 95% CI: 10 . 0-59 . 1-fold ) increase in ZIKV infection reporting after 11/11/2015 . Conversely , assuming that targeted abortions after 11/11/2015 were the only change , 92 . 5% ( mean , 95% CI: 89 . 8-94 . 9% ) of microcephaly-affected births would need to have been aborted to explain the lack of a second peak , corresponding to 1090 ( 803-1480 ) aborted pregnancies between 02/03/2016 and 31/12/2016 . Fig 5D shows how the total number of aborted microcephaly-affected births , which may be observable , would change with different abortion rates of microcephaly-affected births . If microcephaly reporting accuracy were the only factor to change , then a 601% ( mean , 95% CI: 492-726% ) reporting rate of microcephaly cases prior to 13/03/2016 relative to fixed 100% accurate reporting after 13/03/2016 would have been necessary . Accurate data on the true number of abortions in this time period and information on the changes in ZIKV and microcephaly reporting would help to clarify the relative contributions of these mechanisms . Overall , these results highlight the limitations of currently publicly available population-level data in explaining epidemiological trends . Different datasets suggest different risk profiles , some of which contrast with previous population-scale analyses . Whilst data from Bahia , Brazil were suggestive of a risk profile similar to that estimated using data from French Polynesia , data from Colombia and Rio Grande do Norte , Brazil suggest a much longer gestational risk period . [6] Although reporting bias may explain the differences in inferred microcephaly risk in different locations , heterogeneity in the distribution of additional host risk factors of microcephaly may be important . Interpretation of epidemiological data for dengue infection requires an understanding of pre-existing immunity due to the presence of antibody-dependent enhancement , which may also be relevant to the interpretation of CZS incidence given the potential role of dengue antibodies in ZIKV disease enhancement . [54–56] Observations of increased prior dengue exposure in areas of disproportionately increased microcephaly incidence would support this hypothesis and be of importance for dengue- but not yet ZIKV-affected areas , highlighting the need for comprehensive serological studies . [31 , 57] An understanding of other potential host risk factors that may differ between affected areas , such as socioeconomic status or maternal smoking , will further aid the interpretation contrasting incidence data . [32] A limitation of our model is the aggregation of data into high-level administrative units , which may mask small-scale heterogeneity in infection risk and case reporting . This may be particularly problematic in our analysis for Colombia , as using the entire Colombian population and birth numbers as the susceptible population may underestimate the true risk should only a fraction of the population actually be exposed to ZIKV infection . [58 , 59] Similarly , differences in transmission peak times at a small spatial scale coupled with location-specific reporting accuracy may reduce the reliability of the population-wide inferred risk profile . Although we were unable to fit the model at a smaller administrative unit due to the lack of necessary meta-data for Colombia , doing so may reveal a similar risk profile to that estimated using data from Northeast Brazil . Our estimates suggest that ZIKV infection reporting rates would need to have increased 18 . 9-fold ( mean , 95% CI: 10 . 0-59 . 1-fold ) to explain the lack of a second microcephaly wave in Bahia , Brazil on its own , which may have been possible if awareness and diagnostic accuracy improved through the epidemic . We note that syndromic ZIKV reports may have included misclassified CHIKV infections which may not have represented an increased risk of ZIKV-associated microcephaly during the second wave in 2016 . [60] A 18 . 9-fold increase in ZIKV reporting as estimated here could therefore mean that ZIKV reporting was a more accurate representation of the true ZIKV attack rate in 2016 , or that 18 Chikungunya cases were misclassified as ZIKV for every 1 true reported ZIKV case with no change in the proportion of true ZIKV cases that were reported . [41] However , during the period in which second waves of ZIKV infection occurred , there was sufficient virological testing to justify confidence in the relative specificity of reported ZIKV cases . [61] Furthermore , in Salvador , Brazil , where serological data are available , the increase in CHIKV seropositivity from 2015 to 2016 was far lower than for ZIKV seropositivity . [31] Nonetheless , diagnostic tools with improved sensitivity and specificity in distinguishing these infections would help to clarify the proportion of true ZIKV infection incidence that observed incidence data represent . We estimated that 1090 ( mean , 95% CI: 803-1480 ) microcephaly-affected births would need to have been aborted between 02/03/2016 and 31/12/2016 to explain the observed data through increased abortions alone . Given that approximately 1000 abortions are reported in Northeast Brazil weekly , it may be possible to identify the true increase in abortion rate during this time period if and when complete data become available ( Supplementary Material of [11] ) . [35] Estimating the true shape and magnitude of the underlying gestational-age-varying risk profile requires additional data that could either be gathered retrospectively or through surveillance in areas where the first wave of transmission is ongoing or has not yet happened . A key limitation of the epidemiological data gathered in Brazil during 2015 and early 2016 is that surveillance systems were implemented during the epidemic , leading to possible inconsistencies in case definitions and ascertainment rates . Retrospective regional serological surveys have been suggested previously as a means of inferring attack rates , which would constrain estimates for the reporting rate of microcephaly and ZIKV infection and in turn constrain estimates of both the underlying risk and potential changes in behaviour/reporting in the second wave . [60 , 62] In particular , community seroprevalence studies of ZIKV antibodies in women of child-bearing age would provide an accurate estimate of the true proportion of ZIKV-infected women during the outbreak irrespective of symptomatic status and time of infection . In terms of future outbreaks , consistent and accurate case definitions for microcephaly and CZS , —such that sensitivity and specificity are high throughout the epidemic period—would greatly increase the utility of clinical surveillance data for population-level analysis . A key remaining question is whether or not the epidemiological data from Brazil accurately represent the relationship between ZIKV infection and microcephaly , and indeed the wider set of outcomes associated with CZS . Retrospective cohort studies for women of childbearing age to assess whether changes in behaviour regarding conception and infection avoidance occurred in 2016 should clarify whether the second season of ZIKV/microcephaly in Brazil is fully consistent with estimates of gestational-age-varying risk from the first season . [53] If actual reporting rates and behaviour changes are not sufficient to explain the apparent discrepancy between first-wave incidence in Brazil compared to later and elsewhere , the investigation of other potential cofactors , such as prior arbovirus infection , becomes a higher priority . It should then be possible to accurately calculate the risk of CZS based on gestational age at infection and the presence or absence of other possible cofactors . | Zika virus ( ZIKV ) infection is associated with the rise of microcephaly cases observed in Northeast Brazil at the end of 2015 . For women in endemic or at-risk areas , understanding how the relationship between time of infection and microcephaly risk varies through pregnancy is important in informing family planning . However , a relatively modest number of congenital Zika syndrome cases have been observed following subsequent waves of ZIKV infection , limiting our understanding of gestational risk . We used a mathematical model to quantify the shape and magnitude of the gestational-age-varying risk to a fetus . Although the risk profile should be conserved regardless of location , we estimated different profiles when using surveillance data from locations in Northeast Brazil and Colombia . Our results suggest that time-dependent reporting changes likely confound the interpretation of currently available surveillance data . Furthermore , we investigated a range of behavioural and reporting rate changes that could explain two waves of ZIKV infection in Bahia , Brazil despite only one wave of microcephaly . Plausible changes in reporting could explain these data whilst remaining consistent with the hypothesis that ZIKV infection carries a significant risk of microcephaly . Further evidence is needed to disentangle the true risk of congenital Zika syndrome from time-varying reporting changes . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results/Discussion"
]
| [
"medicine",
"and",
"health",
"sciences",
"microcephaly",
"pathology",
"and",
"laboratory",
"medicine",
"maternal",
"health",
"obstetrics",
"and",
"gynecology",
"pathogens",
"geographical",
"locations",
"microbiology",
"vector-borne",
"diseases",
"viruses",
"developmental",
"biology",
"women's",
"health",
"rna",
"viruses",
"morphogenesis",
"birth",
"infectious",
"diseases",
"south",
"america",
"medical",
"microbiology",
"birth",
"defects",
"epidemiology",
"microbial",
"pathogens",
"arboviral",
"infections",
"congenital",
"disorders",
"brazil",
"people",
"and",
"places",
"colombia",
"flaviviruses",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"organisms",
"zika",
"virus"
]
| 2018 | Potential inconsistencies in Zika surveillance data and our understanding of risk during pregnancy |
One of the outstanding challenges in comparative genomics is to interpret the evolutionary importance of regulatory variation between species . Rigorous molecular evolution-based methods to infer evidence for natural selection from expression data are at a premium in the field , and to date , phylogenetic approaches have not been well-suited to address the question in the small sets of taxa profiled in standard surveys of gene expression . We have developed a strategy to infer evolutionary histories from expression profiles by analyzing suites of genes of common function . In a manner conceptually similar to molecular evolution models in which the evolutionary rates of DNA sequence at multiple loci follow a gamma distribution , we modeled expression of the genes of an a priori-defined pathway with rates drawn from an inverse gamma distribution . We then developed a fitting strategy to infer the parameters of this distribution from expression measurements , and to identify gene groups whose expression patterns were consistent with evolutionary constraint or rapid evolution in particular species . Simulations confirmed the power and accuracy of our inference method . As an experimental testbed for our approach , we generated and analyzed transcriptional profiles of four Saccharomyces yeasts . The results revealed pathways with signatures of constrained and accelerated regulatory evolution in individual yeasts and across the phylogeny , highlighting the prevalence of pathway-level expression change during the divergence of yeast species . We anticipate that our pathway-based phylogenetic approach will be of broad utility in the search to understand the evolutionary relevance of regulatory change .
Comparative studies of gene expression across species routinely detect regulatory variation at thousands of loci [1] . Whether and how these expression changes are of evolutionary relevance has become a central question in the field . In landmark cases , experimental dissection of model phenotypes has revealed evidence for adaptive regulatory change at individual genes [2]–[5] . These findings have motivated hypothesis-generating , genome-scale searches for signatures of natural selection on gene regulation . In addition to molecular-evolution analyses of regulatory sequence [6]–[9] , phylogenetic methods have been developed to infer evidence for non-neutral evolutionary change from measurements of gene expression [10]–[12] . Two classic models of continuous character evolution have been used for the latter purpose: Brownian motion models , which can specify lineage-specific rates of evolution on a phylogenetic tree [13]–[16] and have been used to model the neutral evolution of gene expression [11] , [17] , and the Ornstein-Uhlenbeck model , which by describing lineage-specific forces of drift and stabilizing selection [13] , [18] , [19] can be used to test for evolutionary constraint on gene expression [11] , [12] . To date , phylogenetic approaches have had relatively modest power to infer lineage-specific rates or selective optima of gene expression levels . This limitation is due in part to the sparse species coverage typical of transcriptomic surveys , in contrast to studies of organismal traits where observations in hundreds of species can be made to maximize the power of phylogenetic inference [20]–[22] . As a complement to model-based phylogenetic methods , more empirical approaches have also been proposed that detect expression patterns suggestive of non-neutral evolution [23]–[25] . We previously developed a paradigm to detect species changes in selective pressure on the regulation of a pathway , or suite of genes of common function , in the case where multiple independent variants drive expression of pathway genes in the same direction [24] , [26] . Broadly , pathway-level analyses have the potential to uncover evidence for changes in selective pressure on a gene group in the aggregate , when the signal at any one gene may be too weak to emerge from genome-scale scans . However , the currently available tests for directional regulatory evolution are not well suited to cases in which some components of a pathway are activated , and others are down-regulated , in response to selection . In this work , we set out to combine the rigor of phylogenetic methods to reconstruct histories of continuous-character evolution with the power of pathway-level analyses of regulatory change . We reasoned that an integration of these two families of methods could be used to detect cases of pathway regulatory evolution from gene expression data , without assuming a directional model . To this end , we aimed to develop a phylogenetic model of pathway regulatory change that accounted for differences in evolutionary rate between the individual genes of a pathway . We sought to use this model to uncover gene groups whose regulation has undergone accelerated evolution or been subject to evolutionary constraint , over and above the degree expected by drift during species divergence as estimated from genome sequence . As an experimental testbed for our inference strategy , we used the Saccharomyces yeasts . These microbial eukaryotes span an estimated 20 million years of divergence and have available well-established orthologous gene calls [27] , and yeast pathways are well-annotated based on decades of characterization of the model organism S . cerevisiae . We generated a comparative transcriptomic data set across Saccharomycetes by RNA-seq , and we used the data to search for cases of pathway regulatory change .
The Brownian-motion model of expression of a gene predicts a multivariate normal distribution of observed expression levels in the species at the tips of a phylogenetic tree . The variance-covariance matrix of this multivariate normal distribution reflects both the relatedness of the species and the rate of regulatory evolution along each branch of the tree . We sought to apply this model to interpret expression changes in a pre-defined set of genes of common function , which we term a pathway . Our goal was to test for accelerated or constrained regulatory variation in a pathway relative to the expectation from DNA sequence divergence , as specified by a genome tree . To avoid the potential for over-parameterization if the rate of each gene in a pathway were fit separately , we instead developed a formalism , detailed in Methods , to model regulatory evolution in the pathway using a parametric distribution of evolutionary rates across the genes . This strategy parallels well-established models of the rate of DNA sequence evolution across different sites in a locus or genome [28] . Briefly , we assumed that each gene in the pathway draws its rate of evolution from an inverse gamma distribution , and we derived the relationship between the parameters of this distribution and the likelihood of expression observations at the tips of the tree . For each gene , we modeled the contrasts of the expression level in each species relative to an arbitrary species used as a reference , to eliminate the need to estimate the ancestral expression level . A further normalization step , recentering the distribution of expression across pathway genes in each species to a mean of , corrected for the effects of coherent regulatory divergence due to drift . This formalism enabled a maximum-likelihood fit of the parameters describing the pathway expression distribution , given empirical expression data , and could accommodate models of lineage-specific regulatory evolution , in which a particular subtree was described by distinct evolutionary rate parameters relative to the rest of the phylogeny . As a point of comparison , we additionally made use of an Ornstein-Uhlenbeck ( OU ) model [19]: here the rate of regulatory evolution of each gene in a pathway , across the entire phylogeny , was drawn from an inverse-gamma distribution , and all genes of the pathway were subject to the same degree of stabilizing selection , again across the entire tree . Our ultimate application of the method given a set of expression data was to enumerate all possible Brownian motion models in which pathway expression evolved at a distinct rate along the lineages of a subtree relative to the rest of the phylogeny , and for each such model , apply our fitting strategy and tabulate the likelihood of the data under the best-fit parameter set . To compare these likelihoods and the analogous likelihood from the best-fit OU model of universal constraint , we applied a standard Akaike information criterion ( AIC ) [21] , [29] , [30] to identify strongly supported models . As an initial test of our approach , we sought to assess the performance of our phylogenetic inference scheme in the ideal case in which rates of regulatory evolution of the genes of a pathway were simulated from , and thus conformed to , the models of our theoretical treatment . In keeping with our experimental application below which used a comparison of Saccharomyces yeast species as a testbed , we developed a simulation scheme using a molecular clock-calibrated Saccharomyces phylogeny [27] ( see Figure 1a inset ) . We simulated the expression of a multi-gene pathway in which rates of evolution of the member genes were drawn from an inverse gamma distribution . With the simulated expression data in hand from a given generating model , we fit an OU model , an equal-rates model , and models of evolutionary rate shifts in each subtree in turn . Figure 1 shows the results of inferring the mode and rate of evolution from data simulated under a model of accelerated regulatory change on the branch leading to S . paradoxus , and similar results can be seen in Figures S1 through S5 for other rate shift models . As expected , for very small gene groups , inference efforts did not achieve high power or recapitulate model parameters ( Figure 1a , leftmost data point; Figure 1b , leftmost point in each cluster ) , reflecting the challenges of the phylogenetic approach when applied on a gene-by-gene basis to relatively sparse trees like the Saccharomyces species set . By contrast , for pathways of ten genes or more , we observed strong AIC support for the true generating model in cases of lineage-specific regulatory evolution , approaching AIC weights of 100% for the correct model if a pathway contained more than 50 genes ( Figure 1a , Figure S1 and panel a of Figures S2 , S3 , S4 , S5 ) . In these simulations our method also inferred the correct magnitudes of lineage-specific shifts with high confidence , for all but the smallest pathways ( Figure 1b and panel b of Figures S2 , S3 , S4 , S5 ) . Likewise , when applied to simulated expression data generated under models of phylogeny-wide constraint , our method successfully identified OU as the correct model ( Figure 2a ) , though with biased estimates of the magnitude of the constraint parameter when the latter was large ( Figure 2b ) , likely due to a lack of identifiability with the inverse-gamma rate parameter ( Figure S6 ) . We also sought to evaluate the robustness of our method to violations of the underlying model . To explore the effect of our assumption of independence between genes , we simulated a pathway in which expression of the individual genes was coupled to one another and evolving under an equal-rates Brownian motion model , and we inferred evolutionary histories either including or eliminating the mean-centering normalization step of our analysis pipeline . With the latter step in place , our method correctly yielded little support for shifts in evolutionary rates in the simulated data except in the case of extremely tight correlation between genes , a regime unlikely to be biologically relevant ( Figure S7 ) . Additionally , to test the impact of our assumption that the genes of a pathway were all subject to similar evolutionary pressures , we simulated a heterogeneous pathway in which expression of only a fraction of the gene members was subject to a lineage-specific shift in evolutionary rate . Inferring parameters from these data revealed accurate detection of rate shifts even when a large proportion of the genes in the pathway deviated from the rate shift model ( Figure S8 ) . Taken together , our results make clear that the pathway-based phylogenetic approach is highly powered to infer evolutionary histories of gene expression change , particularly lineage-specific evolutionary rate shifts . As a contrast to the poor performance of phylogenetic inference when applied to one or a few genes , our findings underscore the utility of the multi-gene paradigm in identifying candidate cases of evolutionarily relevant expression divergence . We next set out to apply our method for evolutionary reconstruction of regulatory change to experimental measurements of gene expression . The total difference in gene expression between any two species is a consequence of heritable differences that act in cis on the DNA strand of a gene whose expression is measured , and of variants that act in trans , through a soluble factor , to impact gene expression of distal targets . Effects of cis-acting variation can be surveyed on a genomic scale using our previously reported strategy of mapping of RNA-seq reads to the individual alleles of a given gene in a diploid inter-specific hybrid [24] , whereas the joint effects of cis and trans-acting factors can be assessed with standard transcriptional profiling approaches in cultures of purebred species . To apply these experimental paradigms we chose a system of Saccharomyces sensu stricto yeasts . We cultured two biological replicates for each of a series of hybrids formed by the mating of S . cerevisiae to S . paradoxus , S . mikatae , and S . bayanus in turn , as well as homozygotes of each species . We measured total expression in the species homozygotes , and allele-specific expression in the hybrids , of each gene by RNA-seq , using established mapping and normalization procedures , including a variance-stabilizing full-quantile normalization ( see Methods ) . In each set of expression data , we made use of S . cerevisiae as a reference: we normalized expression in the homozygote of a given species , and expression of the allele of a given species in a diploid hybrid , relative to the analogous measurement from S . cerevisiae . To search for evidence of evolutionary constraint and lineage-specific shifts in evolutionary rate in our yeast expression data , we considered as pathways the pre-defined sets of genes of common function from the Gene Ontology ( GO ) process categories . For the genes of each GO term , we used normalized expression measurements in yeast species and , separately , measurements of cis-regulatory variation from interspecific hybrids , as input into our phylogenetic analysis pipeline . Thus , for each of the two classes of expression measurements , for a given GO term we fit models of a lineage-specific rate shift in regulatory evolution incorporating inverse-gamma-distributed rates across genes; an analogous model with no lineage-specific rate shift; and an OU model of universal constraint . The results revealed a range of inferred evolutionary models and AIC support across GO terms ( Figure 3 , Tables 1 and 2 , and Tables S3 and S4 ) , and this complete data set served as the basis for manual inspection of biologically interesting features . Among the inferences of pathway regulatory evolution from our method , we observed many cases of evolutionary interest whose best-fitting model had strong AIC support ( Figure 3 ) . For each of 15 GO terms , cis-regulatory expression variation measurements yielded inference of an evolutionary model with >80% AIC weight ( Figure 3a and Table 1 ) . Many such GO terms represented candidate cases of polygenic regulatory evolution , in which multiple independent variants , at the unlinked genes that make up a pathway , have been maintained in some yeast species in response to a lineage-specific shift in selective pressure on expression of the pathway components . For example , in replicative cell aging genes ( GO term 0001302 ) , cis-regulatory variation measured in interspecific hybrids supported a model of polygenic , accelerated evolution in S . paradoxus ( Figure 4a ) , with some pathway components upregulated and some downregulated in the latter species relative to other yeasts . The total expression levels of cell aging genes in species homozygotes were also consistent with rapid evolution in S . paradoxus ( Figure 4a ) , arguing against a model of compensation between cis- and trans-acting regulatory variation , and highlighting this pathway as a particularly compelling potential case of a lineage-specific change in selective pressure . In other instances , expression measurements in species homozygotes alone supported models of lineage-specific evolution , with each such pathway representing a candidate case of accelerated or constrained evolution at trans-acting regulatory factors . For a total of 41 GO terms , our method inferred models with >80% AIC weight from homozygote species expression data ( Figure 3b and Table 2 ) . These top-scoring pathways included a set of components of the transcription machinery ( GO term 0006351 ) , whose expression levels in S . bayanus were less volatile than those of other yeasts and thus supported a model of lineage-specific constraint ( Figure 4b ) . Additionally , expression of a number of pathways in species homozygotes conformed to the OU model of universal constraint , such as a set of genes annotated in transport ( GO term 0006281 ) , whose expression varied less across all species than would be expected from the genome tree ( Figure 4c ) . Taken together , our findings indicate that evolutionary histories can be inferred with high confidence from experimental measurements of pathway gene expression . In our yeast data , many pathways exhibit expression signatures consistent with non-neutral regulatory evolution , in particular lineages and across the phylogeny . Another emergent trend was the prevalence , across many GO terms , of models of distinct regulatory evolution in the lineage to S . paradoxus as the best fit to expression measurements in species homozygotes ( Figure 3b ) . We noted no such recurrent model in analyses of cis-regulatory variation ( Figure 3a ) , implicating trans-acting variants as the likely source of the regulatory divergence in S . paradoxus . To validate these patterns , we applied our phylogenetic inference method to expression measurements from all genes in the genome analyzed as a single group , rather than to each GO term in turn . When we used expression data from species homozygotes as input for this genome-scale analysis , our method assigned complete AIC support to a model in which the rate of evolution was times faster on the branch leading to S . paradoxus ( AIC weight ) , consistent with results from individual GO terms ( Figure 3b ) . An analogous inference calculation using measurements of cis-regulatory variation , for all genes in the genome , yielded essentially complete support for an OU model of universal constraint ( AIC weight ) . We conclude that constraint on the cis-acting determinants of gene expression , of roughly the same degree in all yeasts , is the general rule from which changes in selective pressure on particular functions may drive deviations in individual pathways . However , for many genes , expression in the S . paradoxus homozygote is distinct from that of other yeasts out of proportion to its sequence divergence , suggestive of derived , trans-acting regulatory variants with pleotropic effects .
The effort to infer evolutionary histories of gene expression change has been a central focus of modern comparative genomics . Against a backdrop of a few landmark successes [11] , [12] , progress in the field has been limited by the relatively weak power of phylogenetic methods when applied , on a gene-by-gene basis , to measurements from small sets of species . In this work , we have met this challenge with a method to infer evolutionary rates of any suite of independently measured continuous characters that can be analyzed together across species . We have derived the mathematical formalism for this model , and we have illustrated the power and accuracy of our approach in simulations . We have generated yeast transcriptional profiles that complement available data sets [31] , [32] by measuring cis-regulatory contributions to species expression differences as well as the total variation between species . With these data , we have demonstrated that our phylogenetic inference method yields robust , interpretable candidate cases of pathway regulatory evolution from experimental measurements . The defining feature of our phylogenetic inference method is that it gains power by jointly leveraging expression measurements of a group of genes , while avoiding a high-dimensional evolutionary model . Rather than requiring an estimate of the evolutionary rate at each gene , our strategy estimates the parameters of a distribution of evolutionary rates across genes . We thus apply the assumption of [10] and model expression of the individual genes of a pathway as independent draws from the same distribution , mirroring the standard assumption of independence across sites in phylogenetic analyses of DNA sequence [33] . Any observation of lineage-specific cis-acting regulatory variation from our approach is of immediate evolutionary interest: a species-specific excess of variants at unlinked loci of common function would be unlikely under neutrality , and would represent a potential signature of positive selection if fixed across individuals of the species . In the study of trans-acting regulatory variation , a priori a case of apparent accelerated evolution of a pathway could be driven by a single mutation of large effect maintained by drift in a species , as in any phenomenological analysis of trait evolution [13] , [34] . Our results indicate that for correlated gene groups , the latter issue can be largely resolved by a simple transformation in which expression of each gene is normalized against the mean of all genes in the pathway . Additional corrections could be required under more complex models of correlation among pathway genes , potentially to be incorporated with matrix-regularization techniques that highlight patterns of correlation in transcriptome data [35] . Similarly , although the assumption of independence across genes could upwardly bias the likelihoods of best-fit models in our inferences , model choice and parameter estimates will still be correct on average even with the scheme implemented here [36] . Our strategy also assumes that the genes of a pre-defined pathway are subject to similar evolutionary pressures . Simulation results indicate that this assumption does not compromise the performance of our method , as we observed robust inference to be the rule rather than the exception even in a quite heterogeneous pathway , if a proportion of the genes evolved under a rate shift model . Although we have used pathways defined by Gene Ontology in this study , our method can easily be applied to gene modules defined on the basis of protein or genetic interactions or coexpression . Any such module is likely to contain both activators and repressors , or other classes of gene function whose expression may be quantitatively tuned in response to selection by alleles with effects of opposite sign [37] , [38] . The phylogenetic approach we have developed here is well-suited to detect these non-directional regulatory patterns , rather than relying on the coherence of up- or down-regulation of pathway genes [24]–[26] , [39]–[41] . Ultimately , a given case of strong signal in our pathway evolution paradigm , when the best-fit model is one of lineage-specific accelerated regulatory evolution , can be explained either as a product of relaxed purifying selection or positive selection on pathway output . Our approach thus serves as a powerful strategy to identify candidates for population-genetic [26] and empirical [40] , [42] tests of the adaptive importance of pathway regulatory change . We have developed an R package , PIGShift ( Polygenic Inverse Gamma rateShift ) , to facilitate the usage of our method . The pathway-level approach is not contingent on the Gaussian models of regulatory evolution we have used here , and future work will evaluate the advantages of compound Poisson process [10] , [43] or more general Lévy process [44] models of gene expression , as well as models that account explicitly for sampling error in expression data . The advent of RNA-seq has enabled expression surveys across non-model species in many taxa . Maximizing the biological value of these data requires methods that evaluate expression variation in the context of sequence divergence between species . As rigorous phylogenetic interpretation of expression data becomes possible , these measurements will take their place beside genome sequences as a rich source of hypotheses , in the search for the molecular basis of evolutionary novelty .
Our basic assumption , following [10] , is that the average expression levels of genes in a pathway evolve as independent replicates of the same Brownian motion or Ornstein-Uhlenbeck process . However , instead of assuming that each gene in the pathway has the same rate of evolution , we allow the different genes in a pathway to draw their rate of evolution from a parametric distribution . As a point of departure , we begin by considering the likelihood of a group of genes whose expression evolves independently , each with its own rate of evolution . Throughout , we use uppercase letters to represent random variables and matrices and lowercase letters to represent nonrandom variables . Assume that we have measured expression of the genes of a pathway in species , and that we have a fixed , time-calibrated phylogeny from genome sequence data describing the relationships between those species . We let be the observations of the expression level of the th gene of the pathway , in each of species . Both the Brownian- motion and Ornstein-Uhlenbeck ( OU ) models predict that the vector is a draw from a multivariate normal distribution with variance-covariance matrix ( where is a scalar—the rate of evolution—and the elements of depend on whether evolution follows the Brownian or Ornstein-Uhlenbeck model; see below ) . Hence , the likelihood of the data is ( 1 ) where is a vector representing the mean expression value at the tips of the phylogenetic tree for gene . Note that where is the th element of . If we assume that there is no branch-specific directionality to evolution , we can avoid the need to estimate in either the Brownian motion model or the OU model by a renormalization of the data . We first arbitrarily choose the gene expression measurements in a single species ( say species 1 ) , and define the new random vector byBy our assumption that there is no branch-specific directionality , so for all and . Because each is multivariate normally distributed with dimension , each will also be multivariate normally distributed with dimension and a slightly different covariance structure . Letting be the covariance matrix corresponding to the , elementary calculations taking into account variances and covariances of sums of random variables reveal thatNext , we wish to incorporate into the Brownian motion and OU models a scheme in which the rates of evolution of the genes of a pathway are not specified independently but instead are drawn from an inverse-gamma distribution . In this context , the genes in a pathway share , the variance-covariance structure due to the tree , but the rate of evolution for each gene is an independent draw from an inverse-gamma distribution . The inverse-gamma distribution has density ( 2 ) where is the gamma function and and are shape and scale parameters . The moments of this distribution areandfrom which it follows that the inverse-gamma distribution has no mean if and no variance if . These properties allow for the distribution of rates of gene expression evolution in a pathway to be relatively broad; in addition , the inverse gamma density has no mass at , which prevents any gene in a pathway from not evolving at all . Also , as and as stays fixed , the distribution converges to a point mass at . Thus , a model where there is one rate for every gene is nested within the inverse-gamma distributed rates model . Computation of the the likelihood of the data under this model is simplified by the fact that the inverse-gamma distribution is the conjugate prior to the variance of a normal distribution . Hence , we see that the likelihood of the observed expression data is ( 3 ) The second line follows recognizing that each integral is independent . Thus , the likelihood of the observations of transcriptome-wide gene expression across the pathway in taxa , normalized by the expression level in taxon , is given by ( 3 ) . For the application to simulated and experimental data as described below , given observations of gene expression of the species at the tips of the tree , and a model that specifies the covariance matrix detailed in the next section , we optimized the log likelihood function using the L-BFGS-B optimization routine in R [45] . In the previous section , we left the unnormalized covariance matrix unspecified . Here we briefly recall the forms of under Brownian motion and the Ornstein-Uhlenbeck process . Define the height of the evolutionary tree to be and and the height of the node containing the common ancestor of taxa and by . Then the covariance matrix for Brownian motion isand the covariance matrix for the Ornstein-Uhlenbeck process iswhere quantifies the strength of stabilizing selection , with large corresponding to stronger selection . To model lineage-specific shifts in the evolutionary rate of gene expression in the context of the Brownian motion model , we adopt a framework similar to that of O'Meara et al . [15] . We assume that in a specified subtree of the total phylogeny , the rate of evolution of every gene is multiplied by a constant , compared to the rest of the tree . Under the Brownian motion model , this is equivalent to multiplying the branch lengths in that part of the tree by that same constant; hence , shifts in evolutionary rate are incorporated by multiplying the branch lengths of affected branches by the value of the rate shift . To evaluate the support for the distinct models we fit to expression data for a given pathway , we require a strategy that will be broadly applicable in cases where no a priori expectation of the correct model is available , such that nested hypothesis testing schemes [15] are not applicable . Instead , given likelihoods from fitting of each model in turn to expression data from the genes of a pathway , we use the Akaike Information Criterion , [46] , to report the strength of the support for each , where is the number of parameters in the model ( for the Brownian motion model in which the rate of evolution is the same along all lineages in the phylogeny , and for all other models ) . For all simulations , we used a phylogenetic tree adapted from [27] by removing the branch leading to Saccharomyces kudriavzevii ( see inset of Figure 1a and Figures S1 , S2 , S3 , S4 , S5 ) . To simulate under models in which each gene in a pathway evolves independently , we generated expression data for one gene at a time as follows . We first drew the rate of evolution from the appropriately parameterized inverse-gamma distribution . Then , without loss of generality , we specified that the expression level at the root of the phylogeny was equal to , and we simulated evolution along the branches of the yeast phylogeny according to either a Brownian motion or an Ornstein-Uhlenbeck process ( with optimal expression level equal to ) , using the terminal expression level on a branch as the initial expression level of its daughter branches . To account for lineage-specific shifts in evolutionary rate in a simulated pathway , we multiplied the rate of evolution of each gene by the rate shift parameter for evolution along the branches affected by the rate shift . For each Brownian motion-based rate shift model applicable to the tree , we simulated 100 replicate datasets for each of a range of gene group sizes , in each case setting , , and the rate shift parameter as specified in Figure 1 and Figures S1 , S2 , S3 , S4 , S5 . For the Ornstein-Uhlenbeck model , we simulated 100 replicate datasets for each of a range of pathway sizes with , , and as specified in Figure 2 . To simulate under models in which expression of genes in a pathway was correlated with coefficient , we first drew , the rate of evolution for each gene , from an inverse-gamma distribution with , . We then parameterized the instantaneous variance-covariance matrix of the -dimensional Brownian motion byso that the distribution of trait change along a lineage was multivariate normal with mean 0 and variance covariance matrix . Separate simulated expression data sets were generated with varying from 0 ( complete independence ) to 1 ( complete dependence ) using 100 replicate simulations for each value . Strains used in this study are listed in Table S1 . For pairwise comparisons of S . cerevisiae and each of S . paradoxus , S . mikatae , and S . bayanus , two biological replicates of each diploid parent species and each interspecific hybrid were grown at 25°C in YPD medium [47] to log phase ( between 0 . 65–0 . 75 OD at 600 nm ) . Total RNA was isolated by the hot acid phenol method [47] and treated with Turbo DNA-free ( Ambion ) according to the manufacturer's instructions . Libraries for a strand-specific RNA-seq protocol on the Illumina sequencing platform , which delineates transcript boundaries by sequencing poly-adenylated transcript ends , were generated as in [48] with the following modifications: 1 ) AmpureXP beads ( Beckman ) were used to clean up enzymatic reactions; 2 ) the gel purification and size-selection step was eliminated; 3 ) the oligo-dT primer used for cDNA synthesis was phosphorothioated at position ten ( TTTTTTTTTT*TTTTTTTTTTVN , V = A , C , G , N = A , C , G , T , * = phosphorothioate linkage , Integrated DNA Technologies ) ; and 4 ) 12 PCR cycles were performed . Libraries were sequenced using 36 bp paired-end modules on an Illumina IIx Genome Analyzer ( Elim Biopharmaceuticals ) . Bioinformatic analyses were conducted in Python and R . RNA-seq reads were stripped of their putative poly-A tails by removing stretches of consecutive Ts flanking the sequenced fragment; reads without at least two such Ts were discarded , as were reads with Ts at both ends . To ensure that expression data from hybrid diploids and purebred species could be compared , for each class of expression measurement for a given pair of species we mapped reads to both species genomes from http://www . saccharomycessensustricto . org [27] using Bowtie [49] with default settings and flags -m1 -×1000 . These settings allowed us to retain only those reads that were unambiguously assigned to one of the two species in each pairwise comparison . A mapped read was inferred to have originated from the plus strand of the genome if its poly-A tail corresponded to a stretch of As at the 3′ end of the fragment , and a read was assigned to the minus strand if its poly-A tail corresponded to a stretch of Ts at the 5′ end of the fragment relative to the reference genome . To filter out cases in which inferred poly-A tails originated from stretches of As or Ts encoded endogenously in the genome , we eliminated from analysis all reads whose stretch of As or Ts contained more than 50% matches to the reference genome . In order to filter out cases of potential oligo-dT mispriming during cDNA synthesis , we also eliminated from analysis all reads that contained 10 or more As in the 20 nucleotides upstream of their transcription termination site . Read mapping statistics can be found in Table S2 . We controlled for read abundance biases due to differing GC content as follows . For each lane of sequencing , we grouped sets of overlapping reads and normalized abundance according to GC content of the overlapping region using full-quantile normalization as implemented in the package EDASeq [50] . Normalized abundance was divided by raw abundance to generate a weight that was assigned to every read in the group . These weights were used in place of raw read counts in all downstream analyses . All expression data are available through the Gene Expression Omnibus under identification number GSE38875 . Coordinates of orthologous open reading frames ( ORFs ) in each genome were taken from http://www . saccharomycessensustricto . org . These ORF boundaries in S . cerevisiae differed , in some cases , from ORF definitions in the Saccharomyces Genome Database [51 , SGD , using the definitions from December 22 , 2007]; genes for which the two sets of definitions did not overlap were discarded . For cases where the definitions overlapped but differed by more than ten base pairs at either end , we used the boundaries defined by SGD and adjusted ortholog boundaries in other species accordingly after performing local multiple alignment [52] of the orthologous regions and flanking sequences as defined by [27] . For most genomic loci , each sense transcript feature was defined as the region from 50 bp upstream to 500 bp downstream of its respective ORF . If sequence within this window for a given target ORF overlapped with the boundaries of an adjacent gene or known non-coding RNA on the same strand , the sense feature boundaries of the target were trimmed to eliminate the overlap . For tandem gene pairs , the 3′ boundary of the upstream gene sense feature was set to 500 bp past the coding stop or the coding start of the downstream gene sense feature , whichever was closer; the 5′ boundary of the downstream gene sense feature was set to 50 bp upstream of its coding start or the 3′ end of the upstream gene sense feature , whichever was closer . We tabulated the GC-normalized expression counts ( see above ) that mapped to each transcript feature for each RNA-seq sample . Given the full set of such counts across all features and all samples , we then applied the upper-quartile between-lane normalization method implemented in EDASeq [50] . The normalized counts from this latter step for a given species were averaged across all biological replicates to yield a final expression level for the feature , which we then transformed and used in all analysis in this work . We downloaded the list of genes associated with each Gene Ontology process term from the Saccharomyces Genome Database and filtered for terms containing at least 10 genes . The resulting set comprised 333 terms . For visual inspection of expression differences between species in Figure 4 , we normalized experimentally measured data by branch lengths ascertained from genome sequence as follows . If expression evolution follows the same Gaussian-based model on all lineages of the yeast phylogeny , when the expression level of gene in taxon is compared to that in taxon used as a reference , the marginal distribution ( the difference in expression between taxon and taxon at gene ) is distributed according to a univariate analog of equation ( 3 ) . In this case , dividing by the absolute branch length according to DNA sequence between taxon and taxon eliminates the dependence of the distribution on the total divergence time between taxa , and the density of this normalized quantity will be the same for all species comparisons . In the case of lineage-specific shifts in evolutionary rate or universal selective constraint , one or more taxa will exhibit distinct densities of the normalized expression divergence measure . Thus , we generated each distribution in Figure 4 by tabulating the log fold-change in expression between the indicated species and S . cerevisiae , and then dividing this quantity by the divergence time between the indicated species and S . cerevisiae according to the genome tree . After this normalization , if a pathway has been subject to accelerated regulatory evolution in one lineage , the distribution of expression log fold-changes corresponding to the species at the tip of that lineage will be wider than expected based on the length of the branch from DNA sequence , and hence it will stand out against the other distributions when plotted as in Figure 4; likewise , constraint on expression evolution of a pathway in a particular species will manifest as a narrower distribution for that species . In the case of a pathway subject to the same degree of regulatory constraint on all branches of the yeast phylogeny , branch lengths ascertained from genome sequence will be large relative to the modest expression divergence , with the most dramatic disparity manifesting when divergent species are compared , yielding the narrowest distribution of normalized expression levels . When visualized as in Figure 4 , the width of the distribution of log fold-changes across genes of the pathway in a given species will thus be inversely proportional to the species distance from S . cerevisiae , with the narrowest distribution for S . bayanus and the widest for S . paradoxus . | Comparative transcriptomic studies routinely identify thousands of genes differentially expressed between species . The central question in the field is whether and how such regulatory changes have been the product of natural selection . Can the signal of evolutionarily relevant expression divergence be detected amid the noise of changes resulting from genetic drift ? Our work develops a theory of gene expression variation among a suite of genes that function together . We derive a formalism that relates empirical observations of expression of pathway genes in divergent species to the underlying strength of natural selection on expression output . We show that fitting this type of model to simulated data accurately recapitulates the parameters used to generate the simulation . We then make experimental measurements of gene expression in a panel of single-celled eukaryotic yeast species . To these data we apply our inference method , and identify pathways with striking evidence for accelerated or constrained regulatory evolution , in particular species and across the phylogeny . Our method provides a key advance over previous approaches in that it maximizes the power of rigorous molecular-evolution analysis of regulatory variation even when data are relatively sparse . As such , the theory and tools we have developed will likely find broad application in the field of comparative genomics . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
]
| []
| 2013 | Inferring Evolutionary Histories of Pathway Regulation from Transcriptional Profiling Data |
To determine the relationship between plasma zinc values and the severity of dengue viral infection ( DVI ) and DVI-caused hepatitis . A prospective cohort study was conducted during 2008–2010 in hospitalized children aged <15 years confirmed with DVI . Complete blood count , aspartate aminotransferase ( AST ) , alanine aminotransferase ( ALT ) , and zinc values ( mcg/dL ) were determined twice: first during the toxic phase ( Zn1 ) and secondly two weeks after recovery ( Zn2 ) . 39 patients were enrolled with a mean age of 9 . 7±3 . 7 years , and 15/39 diagnosed with dengue shock syndrome ( DSS ) . Zn1 values were lower than Zn2 values [median ( IQR ) : 46 . 0 ( 37 . 0 , 58 . 0 ) vs 65 . 0 ( 58 . 0 , 81 . 0 ) mcg/dL , respectively , p <0 . 01] . Zn1 but not Zn2 values had a negative correlation with AST and ALT ( rs = −0 . 33 , p = 0 . 04 and rs = −0 . 31 , p = 0 . 05 , respectively ) . Patients with DSS had lower Zn1 but not Zn2 values compared with non-DSS patients [median ( IQR ) Zn1 , 38 . 0 ( 30 . 0 , 48 . 0 ) vs 52 . 5 ( 41 . 2 , 58 . 7 ) , p = 0 . 02; Zn2 , 61 . 0 ( 56 . 0 , 88 . 0 ) vs 65 . 0 ( 59 . 5 , 77 . 5 ) , respectively , p = 0 . 76] . Zn1 values showed a decreasing trend across increasing dengue severity groups ( p = 0 . 02 ) . Age <5 years and DVI-associated diarrhea were associated with low Zn1 . Children who had a higher grade of dengue disease severity and liver cell injury had lower Zn1 values . Low Zn1 values were probably caused by loss from diarrhea and from zinc translocating to liver cells .
The immunopathogenesis of dengue viral infection ( DVI ) is not well understood , and the level of disease severity is multifactorial , depending on various factors such as viral virulence , secondary DVI , immune response to DVI , and host factors including genetic and nutritional status [1]–[3] . Plasma leakage during the toxic phase of illness , caused by increased endothelial permeability , plays an important role in dengue hemorrhagic fever ( DHF ) /dengue shock syndrome ( DSS ) . Previous studies have found that dengue disease severity was associated with concentrations of pro-inflammatory cytokines and cell apoptosis [4]–[9] . Other studies have found that obese or malnourished children with DVI had higher morbidity and mortality rates than those with normal body weight , suggesting that nutritional status might play an important role in the immunopathogenesis of DVI [3] , [4] , and also that obese and malnourished children had higher proportions of zinc deficiency than normal body weight children [10]–[13] . Zinc deficiency is an important problem in school children , particularly in developing countries [13] . In Thailand , more than half of school children tested had zinc deficiency [14] , [15] . Zinc also functions as an antioxidant and membrane stabilizer . Zinc deficiencies can result in inefficient clearing of infections by impairing innate and adaptive immune responses , creating an imbalance of pro- and anti-inflammatory cytokines , and induction of cell death via apoptosis [16]–[22] . Tumor necrotic factor ( TNF ) has been found to induce zinc deficient-endothelial cells to produce a higher number of inflammatory cytokines than non-zinc deficient endothelial cells , but the production of these inflammatory cytokines was partially inhibited by prior zinc supplementation , suggesting that zinc is a protective and critical nutrient for maintenance of endothelial integrity [23] . The liver is the major target organ of the dengue virus and severity of liver injury is associated with dengue disease severity . A previous study found that zinc supplementation in children who had chronic liver disease could prevent liver injury during treatment with pegylated interferon alpha and ribavirin [24] . These various findings suggest that zinc could play an important role in the immune response to DVI . To our knowledge , there has been only one study involving plasma zinc concentrations collected in the first day of admission in patients with DVI , which did not find any correlation between disease severity and zinc concentrations [25] . However , the findings of one study are not conclusive , as there could be factors that interfere with zinc concentrations at different stages of DVI , meaning the time of sample collection could be important . The current study collected plasma samples during both the toxic phase and then 2 weeks after the patient recovered in order to examine potential correlations between plasma zinc values and dengue disease severity and liver injury .
Permission from the institutional review board of Prince of Songkla University was obtained prior to conducting the study . Parents/guardians provided written , informed consent on behalf of all child participants . Descriptive statistics were used to describe the baseline characteristics of the patients . Comparisons of variables between patients with and without DSS , and with and without severe zinc deficiency , were made using the Mann-Whitney U-test . Fisher's exact test was used for comparison of categorical variables . Zinc levels in the toxic phase were compared graphically with those in the recovery phase and compared statistically using the Wilcoxon -signed rank test and Spearman correlation coefficient . Zinc levels in the toxic phase and in recovery phase were compared across dengue severity groups using a non-parametric test for trends [28] . The correlation of zinc with the AST and ALT levels in the toxic phase were examined using Spearman correlation . A p-value of <0 . 05 was considered statistically significant . All analyses were performed using Stata version 10 ( StataCorp , College Station , Texas ) .
Of the 39 patients admitted during the study period with DVI , 22 ( 56 . 4% ) were male and the mean age was 9 . 7±3 . 7 years ( range 9 months to 14 years ) . Of these , 6/39 ( 15 . 4% ) were obese and none were underweight . The median WSDS was 0 . 2 ( range −1 . 9 to 4 . 1 ) . DF and DHF grades I , II , III , and IV were diagnosed in 7 , 12 , 5 , 13 , and 2 patients , respectively . Primary and secondary DVI were diagnosed in 3 and 36 patients , respectively . Of the 3 patients who had primary DVI , 2 had DHF grade III ( both were infants , aged 9 months and 1 year ) , and the other had DHF grade II ( age 7 . 3 years ) . Nausea or vomiting , upper respiratory symptoms ( cough or runny nose ) and diarrhea were found in 84 . 6% , 20 . 5% , and 23 . 1% of the cases , respectively . Dual infections were found in 3 patients with DSS , one each of urinary tract infection , shigellosis , and scrotal cellulitis . All 3 patients who had dual infection also had diarrhea . None of the patients in the study died from the disease . Five of the DSS patients developed hepatic encephalopathy , while the patient with DHF grade IV with hepatic failure also had respiratory failure and active bleeding . Of the 16 patients who had hemorrhagic symptoms , 4 needed a blood transfusion to control bleeding . The parent/guardian of the four patients did not allow their blood to be sampled during the recovery phase and are therefore omitted from the comparison with blood parameters in the toxic phase . In the remaining patients , the plasma zinc values measured from samples taken during the toxic phase were significantly lower than in blood collected 2 weeks after recovery [median ( IQR ) : 46 . 0 ( 37 . 0 , 58 . 0 ) vs 65 . 0 ( 58 . 0 , 81 . 0 ) mcg/dL , respectively , p<0 . 01] . There was no correlation between plasma zinc values collected during the toxic phase and after recovery ( rs = 0 . 04; p = 0 . 84 ) ( Figure 1 ) . During the toxic phase , all 39 patients except one had zinc values less than 70 mcg/dL; moderately ( 40–60 mcg/dL ) and markedly decreased plasma zinc values ( <40 mcg/dL ) were found in 21 ( 53 . 8% ) and 13 ( 33 . 3% ) patients , respectively . All 3 patients with a dual bacterial infection , all 5 patients with hepatic encephalopathy , and 7/9 patients with acute diarrhea had plasma zinc values <40 mcg/dL . The duration of admission was longer in those who had zinc values lower than 40 mcg/dL than in those who had zinc values higher than 40 mcg/dL [median ( IQR ) : 4 ( 3 , 8 ) vs 3 ( 2 , 4 ) days , p<0 . 01] . The proportions of gender , or patients with respiratory symptoms , nausea or vomiting , obesity , or hemorrhagic symptoms were not different in those who had plasma zinc values lower or higher than 40 mcg/dL . The median plasma zinc values were lower in those younger than 5 years , having diarrhea , dual infection , DSS , or hepatic encephalopathy compared to those who did not have these conditions ( Table 1 ) . The zinc value in the toxic phase showed a significant decreasing trend across increasing dengue disease severity groups in which the median ( IQR ) plasma zinc values in patients with DF , DHF grades I and II , and DSS were 53 . 0 ( 42 . 0 , 60 . 0 ) , 52 . 5 ( 42 . 2 , 58 . 7 ) , 50 . 0 ( 37 . 0 , 62 . 0 ) and 38 . 0 ( 30 . 0 , 48 . 0 ) mcg/dL , respectively , p = 0 . 02 ( Figure 2 ) . When the patients were classified into DSS and non-DSS groups , the plasma zinc values in the DSS group were significantly lower than in the non-DSS group [median ( IQR ) : 38 . 0 ( 30 . 0 , 48 . 0 ) vs 52 . 5 ( 41 . 2 , 58 . 7 ) mcg/dL , respectively , p = 0 . 02] . Two weeks after recovery , the average plasma zinc values had increased in all dengue severity groups . Of the 35 patients who had a blood test in the recovery phase , 20 ( 57 . 1% ) had zinc values lower than 70 mcg/dL . The plasma zinc values in patients with DF ( n = 7 ) and DHF grades I ( n = 12 ) , II ( n = 5 ) , DSS ( n = 11 ) were not significantly different , at medians ( IQR ) 65 . 0 ( 63 . 0 , 78 . 0 ) , 62 . 5 ( 56 . 5 , 71 . 5 ) , 76 . 0 ( 66 . 50 , 83 . 0 ) and 61 . 0 ( 56 . 0 , 88 . 0 ) mcg/dL , respectively , p = 0 . 97 ( Figure 2 ) . Plasma zinc values in those with and without DSS were not significantly different , at medians ( IQR ) 61 . 0 ( 56 . 0 , 88 . 0 ) vs 65 . 0 ( 59 . 5 , 77 . 5 ) , respectively , p = 0 . 76 . Plasma zinc values measured during this period were not correlated with disease severity . Obese and non-obese patients had zinc values <70 mcg/dL in 4/5 ( 80 . 0% ) patients and 17/30 ( 56 . 7% ) patients , respectively . The median ( IQR ) plasma zinc values in obese and non-obese patients were 58 . 0 ( 45 . 5 , 68 . 5 ) and 66 . 5 ( 58 . 7 , 82 . 7 ) mcg/dL , respectively , p = 0 . 07 . During the toxic phase , plasma zinc values had a reverse correlation with both AST and ALT ( rs = −0 . 33 , p = 0 . 04 and rs = −0 . 31 , p = 0 . 05 , respectively ) ( Figure 3 ) . However , the plasma zinc levels had no correlation with albumin and ALP levels , or with total numbers of white blood cells , lymphocytes , or platelets . Two weeks after recovery , all of the LFTs and CBCs had returned to normal values , and the plasma zinc values had no correlation with AST , ALT , albumin , or ALP values , or total numbers of white blood cells , lymphocytes , or platelets .
In this study , most of our patients with DVI had low plasma zinc values when measured during the toxic phase of illness . Patients with DSS had lower plasma zinc concentrations than non-DSS patients , and plasma zinc concentrations had a negative correlation with liver enzymes . Plasma zinc concentrations collected 2 weeks after discharge from the hospital had returned to normal values in half of the patients . The only other study we know of that has examined zinc levels in DVI patients was a study by Widagdo in 2008 . Widagdo found low plasma zinc values ( ≤60 mcg/dL ) in 34/45 ( 75 . 6% ) children with DVI; the plasma zinc values were lowest in patients with DHF grade IV ( plasma zinc values in DHF grades I , II , III and IV were 39 . 9±43 . 2 , 45 . 1±41 . 2 , 80 . 4±31 . 4 , and 7 . 8±1 . 3 mcg/dL , respectively ) [25] . We too found low plasma zinc values ( ≤60 mcg/dL ) in most of our cases ( 87 . 2% ) . However , unlike our study which found that zinc levels in the toxic phase showed a significant decreasing trend across increasing dengue disease severity , the Widagdo study found a discordance between plasma zinc values and disease severity , in which patients with DHF grade III had higher plasma zinc values than those with DHF grades I and II . The Widagdo study did not find any correlations between plasma zinc values and diarrhea as in our study , but found a negative correlation between plasma zinc value and lymphocyte count , which we did not find . These different results between our study and Widagdo could be explained by noting the different times when blood samples were collected for plasma zinc assays; in the Widagdo study , the samples were collected on the first day of admission , but in our study we collected the blood during the toxic phase of illness when the inflammatory cytokines were surging , which is the period during which zinc homeostasis is most likely to be affected during DVI [4]–[9] . We found DVI-associated diarrhea in 9/39 ( 23 . 1% ) patients , which was similar to previous studies which found rates of DVI-associated diarrhea of 17–35% [29] , [30] . One study in adult patients found that patients who had DVI-associated diarrhea were more likely to have more severe DVI compared to those who had no diarrhea [31] . Although we found a higher proportion of DVI-associated diarrhea in patients with DSS vs non-DSS patients ( 33 . 3% vs 16 . 7% , respectively ) , the difference was not significant . DVI-associated diarrhea might be explained by an increased number of inflammatory cytokines , which directly affects and leads to zinc loss through the gastrointestinal tract [32] , which we found to be the most likely factor explaining the decreasing plasma zinc values during the toxic phase of DVI . We found , as in the Widagdo study , that plasma zinc values during DVI did not correlate with nutritional status [25] . We also found that plasma zinc values during the toxic phase did not correlate with poor appetite during illness , but they did correlate with disease severity and liver injury . In addition , 2 weeks after recovery , the average plasma zinc values had markedly increased by 1 . 4 times the plasma values during the toxic phase . These findings suggest that during the toxic phase of DVI , inflammatory response and liver injury cause zinc translocation from the plasma into the liver to prevent oxidative damage to liver tissue , and then after recovery , the zinc translocates again , this time from the liver to the plasma , causing post-illness plasma zinc values to markedly increase compared to toxic phase plasma values [33] , . AST and ALT values in this study had only a moderate negative correlation with zinc concentrations , and therefore there are obviously other important factors accounting for lowered zinc levels during the toxic phase of DVI . Diarrhea appears to be one such factor influencing decreased plasma zinc levels . Although 2 weeks after recovery all of our patients had a normal appetite , and the clinical profiles of illness and LFTs had returned to normal , the plasma zinc values of only half of the patients had returned to normal values ( ≥70 mcg/dL ) . When taken in light of previous studies in Thailand which found that more than half of school children tested had zinc deficiency [14] , [15] , these findings suggest that our patients' baseline plasma zinc values may have been low to begin with , and the most likely cause of zinc deficiency would be a low-zinc diet . Although our study found that obese children had lower zinc values than those who were not obese , there were too few obese patients in the study to attach any significance to this finding . We did not find any correlation of post-illness plasma zinc values and dengue disease severity and liver injury . We also note that plasma zinc values measured 2 weeks after the patients recovered cannot be assumed to represent the pre-illness plasma zinc values and thus speculate about whether pre-illness zinc status may be associated with dengue disease severity . To explore this question , researchers would have to know the pre-illness plasma zinc status of enrolled patients , which would rather impractically involve monitoring the plasma zinc values of a large number of children , on the chance that some of them would later develop DVI . Although in normal humans 75% of the plasma zinc is loosely bound to albumin , and zinc is a component of ALP [35] , we did not find any correlation between values of plasma zinc and albumin or ALP in samples collected either during the toxic phase or 2 weeks after recovery . Taking these factors together suggests that the study patients who had a high inflammatory cytokines response to DVI could have subsequently developed severe DVI , liver injury , and low plasma zinc values from zinc loss , partially from diarrhea and partially from zinc translocating to liver cells . Although our findings are potentially important in considering modifications to current DVI management , we do note that we had a small sample size , especially in regards to the number of patients with DHF grade IV , and future studies with a sufficient sample size are required to further explore our findings and tentative conclusions . | Dengue viral infection ( DVI ) is endemic in tropical counties and severe DVI is a significant cause of death , especially in children . Increased vascular endothelial permeability during the defervescence stage of DVI leading to plasma leakage plays an important role in dengue disease severity . Zinc is a protective and critical nutrient for maintenance of endothelial integrity , and also functions as an antioxidant and membrane stabilizer . Previous studies have found that zinc supplements in children who had diarrhea and sepsis improved the clinical outcomes . Zinc deficiency is common in school children , the age group that commonly acquires DVI , particularly in developing countries . However , prior to studying the potential benefits of zinc supplementation in DVI , having some baseline information concerning plasma zinc values and their correlation with dengue disease severity is necessary . We performed a prospective cohort study during 2008–2010 in 39 hospitalized children aged <15 years confirmed with DVI , and found that plasma zinc values during the toxic phase of disease showed a decreasing trend across increasing dengue severity groups , and also correlated with liver cell injury . DVI-associated diarrhea was probably a major cause of markedly decreased plasma zinc values . These findings will be useful as background information in further studies of whether zinc supplementation can improve the clinical outcome of DVI . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
]
| [
"medicine"
]
| 2013 | Is Zinc Concentration in Toxic Phase Plasma Related to Dengue Severity and Level of Transaminases? |
Production of protein containing lengthy stretches of polyglutamine encoded by multiple repeats of the trinucleotide CAG is a hallmark of Huntington’s disease ( HD ) and of a variety of other inherited degenerative neurological and neuromuscular disorders . Earlier work has shown that interference with production of the transcription elongation protein SUPT4H results in decreased cellular capacity to transcribe mutant huntingtin gene ( Htt ) alleles containing long CAG expansions , but has little effect on expression of genes containing short CAG stretches . zQ175 and R6/2 are genetically engineered mouse strains whose genomes contain human HTT alleles that include greatly expanded CAG repeats and which are used as animal models for HD . Here we show that reduction of SUPT4H expression in brains of zQ175 mice by intracerebroventricular bolus injection of antisense 2’-O-methoxyethyl oligonucleotides ( ASOs ) directed against Supt4h , or in R6/2 mice by deletion of one copy of the Supt4h gene , results in a decrease in mRNA and protein encoded specifically by mutant Htt alleles . We further show that reduction of SUPT4H in mouse brains is associated with decreased HTT protein aggregation , and in R6/2 mice , also with prolonged lifespan and delay of the motor impairment that normally develops in these animals . Our findings support the view that targeting of SUPT4H function may be useful as a therapeutic countermeasure against HD .
Huntington’s disease ( HD ) is one of a collection of untreatable and devastating neurodegenerative and neuromuscular diseases that result from expansion of segments of trinucleotide repeats ( TNRs ) present within certain genes [1–3] . Whereas the huntingtin ( HTT ) gene normally includes fewer than 30 repeats of the glutamine-encoding trinucleotide CAG , expansion to 36 or more repeats results in HTT protein containing a long polyglutamine stretch , leading to HTT protein aggregation and non-canonical protein-protein interactions—and ultimately resulting in neuronal cell death [4–7] . Analogous TNR expansions in other genes underlie certain spinocerebellar atrophies , muscular dystrophies , and other polyglutamine ( polyQ ) -associated disorders [6–8] . Additional diseases are attributable to expansions of other TNRs or to CAG expansions in non-protein-coding regions of other genes [9–12] . Earlier work has shown that the transcription elongation protein SUPT4H ( known in yeast as Spt4 ) , which interacts with its partner SUPT5H ( in yeast , Spt5 ) to form a complex that aids RNA polymerase II processivity [13] , is selectively needed for transcription through gene segments containing expanded TNRs . Decreased production of SUPT4H or Spt4 in cultured cells impedes transcription through expanded TNRs and reduces synthesis of protein containing lengthy polyQ stretches without significantly affecting the production of mRNA and protein from alleles containing non-expanded TNRs . In yeast cells , null mutation of spt4 and consequently , reduced transcription through DNA containing lengthy TNRs , can decrease the abundance of and restore functionality to the resulting protein; in mammalian striatal neurons grown in culture , shRNA directed against Supt4h reduces the production , aggregation , and toxicity of mutant HTT protein [13] . The investigations reported here were aimed at learning whether interference with the actions of SUPT4H would selectively decrease the production of Htt mRNA and protein derived from mutant Htt alleles in whole animal murine models of Huntington’s disease , and if so , whether such a decrease would affect the pathological consequences of TNR expansions . Our findings indicate that decrease in SUPT4H production in cerebral cortex neurons by injection of antisense oligonucleotides ( ASOs ) into the brains of mice expressing a human HTT exon containing expanded CAG repeats [14 , 15] reduces the abundance of mutant Htt mRNA and protein , while having little or no effect on expression of the co-existing normal Htt allele . We further found that downregulation of mutant HTT by deletion of a single Supt4h allele in R6/2 HD mice—which contain a lengthy CAG repeat within a transgenically introduced first exon of the human HTT gene [16]—results in delay of the motor function impairment characteristic of these mice and in prolongation of mouse lifespan .
The discovery that transcription of genes containing expanded repeats of CAG or other trinucleotides located in either protein-coding or transcribed non-coding regions of genes is selectively reduced by interference with the actions of the transcription elongation protein SUPT4H or its yeast counterpart , Spt4 [13] identifies SUPT4H as a potential target for therapies for genetic disorders associated with TNR expansions . In initial experiments to investigate this prospect , we injected 2’-O-methoxyethyl-modified antisense oligonucleotide directed against Supt4h mRNA into the brains of zQ175 mice , which have been engineered to carry a human HTT gene exon that includes expanded TNRs [14 , 15] . The genomes of the adult zQ175 HD mice used in these studies contain an endogenous normal murine Htt allele in addition to the modified one . The anti-sense oligonucleotide ( ASO ) used was shown in preliminary studies to result in ~80% reduction of Supt4h mRNA in the mouse endothelioma cell line bEnd . 3 cells ( ATCC CRL-2299 ) . The procedures we employed ( Materials and Methods ) have been used previously to correct a splicing abnormality in the SMN2 gene in transgenic mice [17] , and were also shown to reduce HTT protein production from both Htt alleles in R6/2 , BACHD and YAC128 mice using ASOs directed against the Htt gene [18] . Analysis of extracts of entire cerebral cortices ( S1 Fig ) or lumbar spinal cords collected from mice receiving ASO directed against Supt4h showed reduction of Supt4h mRNA and protein to 40% or 50% of normal ( Fig 1A , 1B ) . This decrease was accompanied by an approximately 30% decrease from the baseline abundance in untreated zQ175 mouse brains of mutant Htt mRNA and protein , which were produced in lesser amounts than Htt mRNA and protein from genes containing unexpanded TNRs—as has been reported previously [13 , 19] . However , injection of ASO directed against Supt4h did not result in a detectable change in expression of the Htt allele containing an unexpanded TNR , in contrast to the decreased expression in both Htt alleles that resulted from injection under identical conditions of an ASO directed against an Htt gene sequence ( Fig 1C and S2 Fig ) . The selective decrease in mutant Htt expression observed during ASO-mediated targeting of Supt4h in mouse brain and spinal cord tissues contrasts with the non-selectively decreased expression of Htt resulting from similarly performed intracerebroventricular bolus injection of ASO directed against Htt mRNA [18] . To learn about the effects of more widespread and prolonged reduction of Supt4h expression in mice , and also to determine the effects of such reduction on phenotypes characteristic of HD , we first constructed a C57BL6/129-derived mouse strain deleted for Supt4h using conventional genetic knockout approaches ( Fig 2 ) . We obtained mice having a deletion of one Supt4h allele , as confirmed by Southern blot analysis ( Fig 2A ) ; however , mating of such Supt4h+/- animals failed to generate viable offspring having deletions in both Supt4h alleles . Instead , analysis of embryos indicated that homozygous knockout of Supt4h was associated with embryonic lethality at day E7 . 5 ( S1 Table ) . Using quantitative RT-PCR ( qRT-PCR ) to assess Supt4h mRNA abundance in Supt4h+/- mice , we found that Supt4h transcripts in cerebral tissue lysates were decreased to approximately 50% of the abundance observed in Supt4h+/+ littermates ( Fig 2B ) ; consistent with this observation , SUPT4H protein was reduced in the striatal and cortical regions of the brain , as determined by immunohistochemistry staining ( Fig 2C ) and Western blot analysis ( Fig 2D ) . Mice showing this extent of decrease in SUPT4H abundance , which corresponds to the decrease that results in reduced mutant HTT toxicity in cultured striatal neurons [13] , were maintained for 18 months without apparent effects on lifespan or motor function . R6/2 mice , which carry a transgenically introduced first exon of human HTT containing an expanded CAG repeat and which robustly show biochemical and behavior characteristics of HD [20 , 21] , have been used extensively to evaluate events that may affect humans afflicted with HD . To evaluate the effects of perturbed Supt4h expression in these mice , we generated a line of R6/2-derived Supt4h+/- animals ( Fig 3A ) . As was observed for Supt4h+/- mice in the C57BL6/129 strain background , whole brains collected from R6/2 Supt4h+/- animals showed approximately 50% reduction of Supt4h abundance relative to R6/2 Supt4h+/+ animals ( Fig 2 ) . Quantitative RT-PCR using conditions that distinguish between expression of wild-type and mutant Htt alleles indicated that deletion of one Supt4h allele in R6/2 mice was accompanied by a marked reduction in mutant Htt mRNA in brain tissue , whereas mRNA production by the wild-type Htt allele was unaltered by the Supt4h gene deletion ( Fig 3B , 3C ) . Western blotting using an antibody that detects only the mutant form of HTT confirmed that expression of the mutant Htt allele was reduced in zQ175 mice treated with ASO directed against either Supt4h or Htt; however , ASO against Htt also reduces protein produced by the normal Htt allele , while ASO directed against Supt4h did not ( S2 Fig ) . In R6/2 mouse experiments , slot blot assays and antibody that detects only the mutant form confirmed the ability of a null mutation in one Supt4h allele to reduce expression of mutant HTT in Supt4h knockout mice as shown in Fig 3B , 3C . Aggregation of mutant HTT is a prominent feature of HD during disease progression , and reduction of such aggregation has been reported to rescue neurons from dysfunction and cell death [22–24] . Our earlier studies using cultured cells showed that both the production and aggregation of mutant HTT is decreased by siRNA directed against Supt4h [13] . We observed that R6/2 mice deleted for one Supt4h allele showed a similarly reduced abundance of mutant HTT protein and a decrease in HTT protein aggregates ( Fig 4B , 4C , and S3 Fig ) , while showing no change in the amount of normal HTT protein synthesized from a coding sequence containing a short TNR ( Fig 4A ) . Additionally , reduction of the DARPP-32 protein , which is highly enriched in medium-sized spiny neurons and has been reported to be down-regulated concurrently with early neuronal dysfunction in the R6/2 mouse model of HD [25 , 26] , was partially reversed in mouse brains by deletion of one Supt4h allele ( Fig 4D ) . Typically , R6/2 mice show severe impairment of motor coordination by 8–12 weeks of age [27] , and die between 13 and 16 weeks of age [16] . The progressive deterioration in motor function can be detected by reduction in the length of time that the mice can remain on a rotating rod—the so-called “rotarod assay” [27] . We employed rotarod performance assays to compare the motor function of R6/2 mice deleted in one Supt4h allele with that of R6/2 Supt4h+/+ littermates . As observed previously [28] R6/2 Supt4h+/+ mice showed a progressive decline in motor function starting at 10 weeks of age; however , in R6/2 Supt4h+/- mice , the decline was not apparent until 13 weeks of age ( Fig 5A ) —suggesting that reduction of SUPT4H abundance by half in these animals , as indicated above , is sufficient to yield measurable benefits in motor function . Similarly , R6/2 mice having one Supt4h allele deleted showed better performance in a beam walking test ( Fig 5B ) commonly used as another parameter of motor function in the HD mouse model system [27 , 28] . R6/2 mice carrying a heterozygous deletion in Supt4h also showed a longer lifespan ( Fig 5C ) than did Supt4h+/+ animals; however , no detectable effect of an Supt4h deletion on the loss of body weight that is characteristic of HD progression in R6/2 mice ( Fig 5D ) was observed .
The results reported here demonstrate that experimentally induced decrease of the transcription elongation protein SUPT4H in brain and spinal cord tissues of murine models of Huntington’s disease results in selectively decreased expression of mutant huntingtin alleles , and that these events are associated with reduction of HTT protein aggregates , delay in the impairment of motor function seen in R6/2 HD mice as the animals age , and an increase in the R6/2 mouse lifespan . SUPT4H and its yeast counterpart , Spt4 , function in cells by binding to Spt5/SUPT5H to form a protein complex; the N-terminal region of SUPT5H then interacts with the C-terminal region of RNA polymerase II ( Pol II ) , an event that is thought to tighten the Pol II clamp around DNA templates and limit dissociation of Pol II from DNA during transcription pauses [13 , 29–31] . Data obtained by crystallographic analyses indicate that SUPT4H and Spt4 do not directly contact the polymerase [32–35] and in yeast , null mutations of spt4 , unlike those of spt5 do not preclude cell viability [36] . Our earlier findings that dissociation of the Pol II complex from DNA in template segments containing expanded TNRs is increased during Spt4 deficiency , and that interference with the function of Spt4 or SUPT4H decreases expression of genes containing expanded TNR regions in cultured cells—while not significantly affecting transcription of genes containing shorter TNRs or no TNRs at all [13]—have raised the prospect that targeting the function of SUPT4H may be a useful strategy for treatment of HD and possibly other TNR diseases . The murine results reported here support this notion . The potential therapeutic value of reducing HTT expression in the brains of individuals afflicted with HD or other polyQ disorders is well recognized [37] , and antisense oligonucleotides that target Htt sequences common to mutant and normal alleles have been shown to reduce overall production of Htt mRNA and protein in brain tissue when delivered into the cerebrospinal fluid of HD-afflicted mice by transient infusion [18] . Such non-selective decrease in expression of both Htt alleles prevented the appearance of HD-disease symptom , and did not result in observable detrimental effects during the duration of those experiments . However , bi-allelic Htt inactivation in the forebrain and testes leads to progressive neuronal degeneration and sterility [38] , and selective targeting expression from mutant HTT alleles has been a desirable therapeutic objective . Selective knockdown of mutant HTT mRNA translation recently has been reported using single-stranded RNA ( ssRNA ) that target expanded CAG repeat segments within HTT transcripts [39]; while this approach reduces mutant HTT protein , it does not affect the abundance of mutant HTT mRNA , which can also contribute to cellular toxicity [40–42] . siRNA and ASOs that target polymorphic gene sequences that differ in mutant and normal HTT alleles have also been reported to achieve allele-specific inhibition [43–46] . However , downregulation of mutant HTT mRNA by the targeting of SUPT4H is a strategy that is independent of fortuitously occurring sequence differences in mutant and wild-type HTT alleles and additionally may also be applicable toward the treatment of other disorders caused by TNR expansions . Even in the presence of SUPT4H , mRNA produced by an Htt allele containing expanded TNRs is less abundant than mRNA from a co-existing allele having unexpanded TNRs [13 , 19 , 47] . The expanded polyQ protein encoded by the mutant allele is correspondingly less abundant [13 , 19] , although the upregulation of translation mediated by increased binding of the MID-1 protein to expanded CAG repeats can elevate mutant HTT protein above the level of wild-type HTT [47] . Reduction of SUPT4H by half in the brains of zQ175 mice by intracerebroventricular bolus injection of ASO or in R6/2 mice by deletion of one Supt4h allele results in a decrease in mutant Htt transcription beyond the already reduced level of mRNA production . In the R6/2 strain , which displays phenotypic features seen in human HD , such reduction resulting from deletion of a single Supt4h allele was associated with partial reversal of the HD-like phenotypic properties . The non-HD mouse strain having an Supt4h allele deleted showed no overt functional impairment during an 18-month period of observation . Our results show that knockdown of Supt4h in murine tissues to 30–50% of normal does not preclude survival of mice after birth . However , notwithstanding the viability of yeast carrying null mutations in the Supt4h ortholog SPT4 [13] , and the minimal effect of shRNA knockdown of Supt4h on RNA-SEQ profiles in mice [13] , the lethality we observed for Supt4h-/- embryos argues that one or more actions of SUPT4H may be required for adequate transcription during embryogenesis of one or more of multiple normal mouse genes that contain >40 trinucleotide repeats . Mutant alleles in most HD patients do not contain repeats of the length necessary to yield HD-related phenotypes in transgenic mice [48 , 49]; and additionally , the consequences of Supt4h knockdown potentially may be affected by genetic variation in the native cellular abundance of SUPT4H ( and perhaps of SUPT5H its transcription elongation complex partner ) in brain tissues of different individuals . While the investigations reported here indicate that the allele-specific effects of Supt4h knockdown reported previously in cultured cells occur also in mouse HD model systems and that reduction of Supt4h expression can result in disease-related consequences in a mouse HD model , the parameters that affect selective expression of mutant vs . wild type HTT alleles require further investigation before the clinical relevance of our findings can be established .
All animal experiments were performed in accordance with the guidelines established by the Institutional Animal Care and Use Committee ( IACUC ) of Academia Sinica . All the experimental protocols were approved by IACUC and the approval number is 11–12–253 . Mice were sacrificed by CO2 inhalation according to the approved protocol for tissue collection and IHC analysis . The zQ175 mouse strain , which carries a normal murine Htt allele and a knock-in ( KI ) mutant Htt/HTT mouse/human hybrid allele containing around 188 CAG repeats [15] , was provided by the CHDI Foundation , Inc . Male R6/2 [B6CBA-Tg ( HDexon1 ) 62Gpb/1J] [16] mice , which contain an N-terminally-truncated mutant HTT allele containing a long CAG repeat , were obtained from Jackson Laboratories ( Bar Harbor , ME , USA ) and mated to normal females of mouse strain B6CBAFI/J . The genotype of offspring was verified by polymerase chain reaction ( PCR ) , using genomic DNA extracted from tail tips and primers that specifically target the mutant Htt transgene . The number of CAG repeats of R6/2 mice used in this study was 240 ± 10 ( mean ± SEM ) . Supt4h knockout was generated in a C57BL6/129S6 hybrid mouse line background using a conventional gene targeting approach . The colony was maintained by breeding Supt4h+/- males with C57BL6 females . PCR genotyping was carried out using primer sets Supt4h WT and Supt4h MT to detect intact and genetically deleted alleles of Supt4h respectively . The nucleotide sequence of these primers is shown in S2 Table . To produce R6/2 mice that contain or lack a deletion of one Supt4h allele , R6/2 males were crossed with Supt4h+/- females , and progeny were subjected to genotyping for both Supt4h intact and genetically deleted alleles and for the R6/2 human Htt transgene containing an expanded CAG repeat . The biochemical and behavioral experiments were performed using littermates from the same population . Mice were housed at the Institute of Biomedical Sciences Animal Care Facility ( Taipei , Taiwan ) under a 12h light-dark cycle . All procedures were accomplished using a protocol approved by the Academia Sinica Institutional Animal Care and Utilization Committee ( Taipei , Taiwan ) . The Supt4h ASO ( 5’-CGACACTTGTGTCCCCTGCT-3’ ) used in this study was a 20-mer oligonucleotide that contains a phosphorothioate backbone and a chimeric 2’-O-methoxyethyl ( MOE ) /DNA design [50] containing five MOE-modified nucleotides at each end of a centered stretch of ten DNAs . Oligonucleotide was synthesized [17] and solubilized in PBS . zQ175 KI mice [15] were kindly provided by CHDI and received ASO ( 300 μg ) or PBS via intracerebroventricular ( ICV ) bolus injections at the age of 5 . 5 months . Tissues were collected 4 weeks after a single ICV bolus injection , and RNA or protein was extracted as previously described [17] . Total RNA was extracted from isolated tissues using Trizol reagent ( Invitrogen ) and the abundance of Supt4h and Htt transcripts was assessed by quantitative real time RT-PCR ( qRT-PCR ) as described previously [13] . For samples collected from zQ175 HD mice , 1 μg of total RNA was converted to cDNA , followed by qRT-PCR analysis using ABI StepOnePlus Real-Time PCR System ( Life Technologies ) . In zQ175 HD mice , the KI mutant allele contains a DNA fragment of human HTT , which is distinct from murine Htt in nucleotide sequence . PCR primers were designed to correspond to species-specific sequences and thus to differentiate mRNA produced from the wild-type vs . KI mutant allele . Samples collected from R6/2 experiments were analyzed as described above , except that 3 μg of total RNA was used for the synthesis of cDNA , and qRT-PCR was performed using ABI PRISM 7500 Sequence Detection System ( Life Technologies ) . Relative gene expression was determined by the 2-△△Ct method after normalization with either U6 or 18S ribosomal RNA . Oligonucleotides used for qPCR are summarized and shown in S2 Table . Genomic DNA extracted from the tails of mice was digested by restriction enzymes Bgl II or BstB l ( New England BioLabs ) . DNA was then electrophoresed on agarose gels , transferred to Hybond-N+ nylon membranes ( GE Healthcare ) , and fixed on membranes using UV cross-linker ( UV Stratalinker 1800 , Stratagene ) , as previously described [51] . DNA probes for detection of the Supt4h locus were generated by PCR and labeled with 32P using Amersham Rediprime II DNA Labeling System ( GE Healthcare ) . After hybridization with Supt4h 5’- or 3’-probe in Church buffer ( 0 . 25 M sodium phosphate , 1 mM EDTA , 1% BSA , 7% SDS , and 10 mg/ml salmon sperm DNA ) overnight , the membrane was rinsed twice with buffer I ( 2X SSC , 0 . 1% SDS ) at 30°C for 30 minutes , followed by buffer II ( 0 . 2X SSC , 0 . 1% SDS ) at 60°C . DNA fragments recognized by the probes were monitored by Typhoon 9410 Variable Mode Imager ( GE Healthcare ) . Plasmid construct pPAL7-HA-Supt4h , which expresses full-length murine SUPT4H , was created by PCR amplification of a DNA fragment encoding SUPT4H and the HA-epitope and subsequently sub-cloning of this PCR product in E . coli on expression vector pPAL7 . The expression construct was introduced into E . coli BL21 ( DE3 ) , and production of SUPT4H protein was induced by isopropyl-β-D-thiogalactopyranoside ( IPTG , Promega ) as per the manufacturer’s protocol . For protein purification , BL21 cells were lysed using a Microfluidizer ( Microfluidics Corp . ) in buffer A ( 0 . 1 M sodium phosphate , pH 7 . 2 ) . After centrifugation , the supernatant was mixed with 1 ml Profinity eXact purification resin ( Bio-Rad ) and incubated at room temperature for 2 hours . Reaction mixtures were loaded onto a Poly-Prep Chromatography Column ( Bio-Rad ) , washed with 10 column volumes of wash buffer , and incubated in 2 column volumes of elution buffer ( 100 mM sodium phosphate , 100 mM sodium fluoride , pH 7 . 2 ) at 4°C overnight . The purified protein was subsequently eluted , transferred to a dialysis membrane ( Cellu‧Sep T1 , Uptima ) , and sent to LTK BioLaboratories for immunization of rabbits . Antibody against SUPT4H was validated by Western blotting using purified recombinant protein , protein lysates of mammalian 293T cells expressing ectopic SUPT4H , and brain lysates obtained from C57BL6 mice . Tissues collected from zQ175 HD mice were homogenized by dounce homogenizer using cell lysis buffer ( Cell Signaling ) . For R6/2 or Supt4h genetically modified mice , brain lysates were prepared similarly using lysis buffer ( 10 mM HEPES , 1 mM DTT , 200 μM Na3VO4 , 8 . 5% ( w/v ) sucrose , protease inhibitor ) . Immunoblotting was performed as previously described [13] . In brief , equal amounts of protein were resolved by electrophoresis on 8 , 12 , or 15% sodium dodecyl sulfate ( SDS ) -polyacrylamide gels , transferred onto immobilon-P polyvinylidene difluoride ( PVDF ) membranes ( Millipore ) , and probed with anti-SUPT4H , anti-α-Tubulin ( DM1A , Sigma ) , anti-TBP ( T1827 , Sigma ) , anti-HTT ( MAB2166 , Chemicon ) , anti-polyQ ( clone 5TF1–1C2 , MAB1574 , Chemicon ) , anti-β-actin ( GTX109639 , GeneTex ) , anti-GAPDH ( GTX100118 , GeneTex ) , or anti-DARPP-32 ( #2302 , Cell Signaling ) antibodies . After incubation with a horseradish peroxidase ( HRP ) -conjugated secondary antibody for 1 h , the immunoreactive signals were detected by ECL reagent ( enhanced chemiluminescence , PerkinElmer ) . Filter-retardation assays were performed as previously described [13] . Briefly , brain lysates collected from R6/2-derived animals were loaded through a slot-blot manifold ( Bio-Rad ) onto CA membranes ( cellulose acetate , 0 . 2 μm pore size; Schleicher & Schuell ) , which retain SDS-insoluble protein aggregates . Membranes were blocked with 5% skim milk in TBST ( 10 mM Tris-HCl , pH 8 . 0 , 150 mM NaCl , 0 . 05% Tween 20 ) and probed with EM48 antibody ( MAB5374 , Chemicon ) at 4°C overnight . EM48 antibody identifies N-terminal huntingtin fragments containing a long stretch of polyglutamine , and is particularly efficient for detecting human huntingtin aggregates , whereas the antibody has only weak affinity for rodent HTT protein [52] . After incubation with the corresponding secondary antibody , immunoreactive signals were detected by ECL reagent and recorded using Fuji X-ray film . Animals were anesthetized before perfusion with 4% paraformaldehyde in PBS ( pH 7 . 4 ) . Brains were removed , post-fixed with 4% paraformaldehyde at 4°C overnight , and embedded in paraffin as previously described [53] . Serial coronal sections ( 5 μm ) were deparaffinized by xylene substitute ( Fluka ) and rehydrated by serial alcohol dilution and subsequent PBS rinse . After heating and cooling in retrieval solution ( pH 6 . 0 , DakoCytomation ) , brain sections were permeabilized by 0 . 5% Triton X-100 and blocked with 10% goat serum for 1 h . Sections were then stained with primary antibodies against SUPT4H or ubiquitin ( DakoCytomation ) at 4°C overnight , followed by incubation with the corresponding secondary antibody for 1 h . To enhance the signal , Vectastain ABC kit ( Vector Laboratories ) was applied before staining with diaminobenzidine ( DakoCytomation ) . Nuclei were stained with hematoxylin or methyl green . Body weights of mice were recorded weekly . Motor coordination was assessed using a rotarod apparatus ( MK-660D , Muromachi-Kikai ) retaining at a constant speed ( 12 rpm ) over a period of 2 minutes , as previously described [28] . Animals were pre-trained at the age of 7 weeks to become acquainted with the apparatus . Then mice were tested three times per week from the age of 8 to 14 weeks . Beam walk analysis was applied to assess motor coordination [28] . Mice were trained to traverse a circular beam having a diameter of 17 mm , followed by testing on an 11-mm-diameter beam once per week . Results were recorded as the duration of time ( Latency ) spent by mice to walk across the 80-cm-long beam . Latency was recorded as 120 seconds when mice spent more than 120 seconds traversing the beam . Values shown in the figures are presented as mean ± SEM . All statistical analyses were carried out by Student’s t-test except indicated otherwise . Rotarod performance , beam walk test , and change of body weight were analyzed using two-way analysis of variance ( ANOVA ) , followed by a post-hoc Bonferroni multiple comparison test . Survival statistics were performed using Log-rank test . All tests were performed using the SigmaPlot software , version 10 . 0 . A value of p<0 . 05 was considered statistically significant . | Huntington’s disease ( HD ) is an inherited genetic disorder that leads to degeneration of brain cells and consequently to abnormal body movements , decreased mental capacity , and death . It is one of a group of untreatable degenerative neurological and neuromuscular diseases caused by expansion of gene segments containing multiple tandemly arrayed copies of short DNA sequences called trinucleotide repeats ( TNRs ) . We report here that interference with production of a protein , SUPT4H , that is differentially needed for transcription through mutant Htt genes containing expanded TNRs reduces synthesis of abnormal Htt messenger RNA and protein , decreases HTT aggregates in murine brains , delays the occurrence of pathological features of HD , and prolongs HD mouse lifespan . Our results suggest that targeting of SUPT4H may be of value in the treatment of HD . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| []
| 2015 | Effects on Murine Behavior and Lifespan of Selectively Decreasing Expression of Mutant Huntingtin Allele by Supt4h Knockdown |
Orderly chromosome segregation during the first meiotic division requires meiotic recombination to form crossovers between homologous chromosomes ( homologues ) . Members of the minichromosome maintenance ( MCM ) helicase family have been implicated in meiotic recombination . In addition , they have roles in initiation of DNA replication , DNA mismatch repair and mitotic DNA double-strand break repair . Here , we addressed the function of MCMDC2 , an atypical yet conserved MCM protein , whose function in vertebrates has not been reported . While we did not find an important role for MCMDC2 in mitotically dividing cells , our work revealed that MCMDC2 is essential for fertility in both sexes due to a crucial function in meiotic recombination . Meiotic recombination begins with the introduction of DNA double-strand breaks into the genome . DNA ends at break sites are resected . The resultant 3-prime single-stranded DNA overhangs recruit RAD51 and DMC1 recombinases that promote the invasion of homologous duplex DNAs by the resected DNA ends . Multiple strand invasions on each chromosome promote the alignment of homologous chromosomes , which is a prerequisite for inter-homologue crossover formation during meiosis . We found that although DNA ends at break sites were evidently resected , and they recruited RAD51 and DMC1 recombinases , these recombinases were ineffective in promoting alignment of homologous chromosomes in the absence of MCMDC2 . Consequently , RAD51 and DMC1 foci , which are thought to mark early recombination intermediates , were abnormally persistent in Mcmdc2-/- meiocytes . Importantly , the strand invasion stabilizing MSH4 protein , which marks more advanced recombination intermediates , did not efficiently form foci in Mcmdc2-/- meiocytes . Thus , our work suggests that MCMDC2 plays an important role in either the formation , or the stabilization , of DNA strand invasion events that promote homologue alignment and provide the basis for inter-homologue crossover formation during meiotic recombination .
Chromosome segregation during the first meiotic division uniquely differs from chromosome segregation during mitosis and the second meiotic division [1 , 2] . Centromeres belonging to sister chromatids are pulled toward opposite spindle poles during mitosis and the second meiotic division . In contrast , centromeres belonging to homologous chromosomes ( homologues ) that originate from different parents are pulled to opposite spindle poles during the first meiotic division . This bi-orientation of homologue centromeres requires homologues to pair and become physically linked before segregation [1 , 2] . In most organisms including mammals , inter-homologue physical linkages are provided by the collaborative action of sister chromatid cohesion and inter-homologue crossovers , the latter of which are formed by meiotic recombination during the first meiotic prophase . Meiotic recombination initiates with the programmed generation of large numbers of DNA double-strand breaks ( DSBs ) ( 200–400 per cell in mice and humans ) by the SPO11 enzyme [3–7] . This results in SPO11-bound DNA ends at break sites [3 , 4] , which are processed to remove SPO11 from DNA –ends and to produce single-stranded 3´ DNA overhangs [8] . These single-stranded DNA ends attract RecA-like recombinases DMC1 and RAD51 , which form “recombinosome” complexes that promote invasion of single-stranded DNA ends into homologous DNA sequences to produce so called displacement-loops ( D-loops ) [9–11] . It is thought that stable strand invasions preferentially occur into homologues as opposed to sister chromatids during meiosis [12–14] . This inter-homologue bias in the formation of recombination intermediates is thought to ensure that DSBs efficiently promote the recognition and the pairing of homologues based on sequence similarity . DNA breaks are formed and become repaired within the context of chromosome axes , which are linear proteinaceous chromatin structures that form along cores of chromosomes during meiosis [15–18] . Upon successful homologue pairing , axes of homologues closely align and get incorporated into a meiosis-specific chromatin structure , called the synaptonemal complex . The synaptonemal complex consists of two parallel axes and transverse filaments that connect the axes to a shared central linear protein structure , called the central element [19] . The synaptonemal complex is thought to signal the end of the homologue pairing process [20–22] and promote the repair of DSBs by recombination [16 , 23–28] . This repair involves DNA synthesis that starts from the 3´end of the invading strands , and uses the invaded homologous sequence as a template [11] . Meiotic recombination-mediated DSB repair has two main pathways with distinct outcomes: reciprocal recombination/crossovers and non-reciprocal recombination/non-crossovers [5 , 11] . At least one strand invasion on each chromosome is thought to be specially stabilized and turned into a crossover . In contrast , most of the strand invasions are repaired as non-crossovers , which often manifest as gene conversions after the completion of repair . Correct homologue pairing and crossover formation require finely balanced activities that either stabilize or destabilize strand invasions and resultant recombination intermediates . The BLM helicase has been suggested to destabilize strand invasion intermediates , and this function might be important for error correction of strand invasions and the dissolution of difficult-to-repair recombination intermediates [29–34] . The strand invasion intermediate destabilizing activity of BLM is counteracted by the MutSɣ complex [31 , 35] , which consists of a heterodimer of MSH4 and MSH5 that form clamps around DNA strand invasion intermediates thereby stabilizing them [36] . Accordingly , MSH4 and MSH5 proteins are necessary for the alignment of homologues , homologous synaptonemal complex formation and the efficient completion of DNA repair during meiosis in mammals [37–39] . Putative helicases of the minichromosome maintenance ( MCM ) protein family have also been implicated in promoting recombination , although MCM proteins were initially discovered as hexameric helicases that are required for the initiation of DNA replication ( reviewed in [40] ) . In particular , three MCM-related Drosophila proteins , REC , MEI-217 and MEI-218 , form a complex and promote meiotic crossover formation by stabilizing strand invasion intermediates , opposing BLM function and inhibiting the non-homologous end joining repair pathway of DNA break repair [41–43] . Although these proteins are not homologous to MSH4 or MSH5 , it was proposed that REC , MEI-217 and MEI-218 substitute for the MutSɣ complex , which is missing from Drosophila [43] . MCM8 , an orthologue of REC , plays important roles in homologous recombination in plants and vertebrates , where the MutSɣ complex is present [44–48] . However , unlike Drosophila REC , vertebrate MCM8 is also important for mitotic recombination and DSB repair [45–48] . Mammalian MCM8 forms a complex with MCM9 in mitotic cells [45 , 48] and is important for resection of DNA ends at break sites at the initial stages of homologous recombination in mitotic cells [47] . In contrast to MCM8 , MCM9 does not play an essential role in meiotic recombination [45] . Curiously , MCM8 is apparently not needed for resection of DNA ends at break sites in meiosis , yet it is important for an as yet undefined recombination step that is essential for efficient homologue alignment and synaptonemal complex formation [45] . This suggests that mammalian MCM8 performs MCM9-independent functions in meiosis , and that like Drosophila REC [43] , mammalian MCM8 might also have a function in stabilizing DNA strand invasion intermediates in meiosis . Interestingly , the REC interacting MEI-217 and MEI-218 proteins of Drosophila also have a predicted orthologue in mammals , called MCMDC2 [43] . Yet , it has not been reported if mammalian MCMDC2 is involved in meiotic recombination . Here we describe the functional analysis of Mcmdc2-/- mice and show that mouse MCMDC2 is crucial for meiotic recombination and DSB repair . More specifically , we hypothesize that MCMDC2 promotes the formation and/or the stabilization of strand invasion intermediates that permit alignment of homologues .
To address if MCMDC2 could play a role in meiotic recombination we asked if Mcmdc2 transcripts are present in testis . Thus , we used RT-PCR to assess expression levels of Mcmdc2 in 17 somatic tissues and testes of mice ( Fig 1a ) . The RT-PCR analysis indicated that Mcmdc2 transcripts were indeed enriched in testis as compared to somatic tissues ( Fig 1a ) . Furthermore , analysis of public databases ( http://www . germonline . org/Homo_sapiens/geneview ? gene=ENSG00000178460 ) [49] showed that human Mcmdc2 was preferentially expressed in the testis , particularly in spermatocytes . Thus , its expression suggested a role for Mcmdc2 in meiosis . Given that the Drosophila homologs of MCMDC2 are required for crossover formation [50–52] , we speculated that mammalian MCMDC2 may also function in meiotic recombination . To test this hypothesis , we attempted to generate antibodies against distinct fragments of mouse MCMDC2 both in rabbit and guinea pig , however none of our antibodies reliably detected MCMDC2 , which precluded localization studies of MCMDC2 . To directly test the biological functions of MCMDC2 we generated mice where Mcmdc2 was disrupted after the 4th exon ( Fig 1b–1d ) . The targeting strategy was designed to terminate the 681-amino acid-long MCMDC2 protein after the 95th amino acid ( Fig 1b ) . This was due to the combined effects of the removal of the 5-7th exons ( encodes 96–237 amino acids of MCMDC2 ) causing a frameshift , and the insertion of a strong ectopic splice acceptor site and a transcriptional terminator into the 4th intron . RT- PCR analysis confirmed strongly reduced expression of Mcmdc2 exons beyond the 4th exon ( including exon 8–11 , which are not deleted from the genome ) in testes of the Mcmdc2-/- mice ( Fig 1d ) . Even transcripts of exon 3–4 , which are upstream of the deletion , were detected at a lower level in Mcmdc2-/- testes than in wild-type testes . MCMDC2 protein fragments that may be produced from these residual transcripts are unlikely to be functional . This is because deletion of the 5-7th exons would allow only a short 95 amino acid N-terminal fragment to be produced from exons 1–4 . Even if alternative splicing generated rare transcripts where sequences from the 4th exon were linked to sequences downstream of the deleted 5-7th exons , protein products of these transcripts would lack most parts of MCMDC2 , including the entire conserved MCM-like region ( SMART: SM00350 , amino acids 177–623 of MCMDC2 ) , because these transcripts would be subject to a frameshift mutation . Mcmdc2-/- mice were viable and did not show any obvious somatic defects . Although the previously published Mcm8-/- and Mcm9-/- mice were viable , Mcm8-/- and Mcm9-/- mouse embryonic fibroblasts ( MEFs ) displayed slow growth and sensitivity to the DNA replication inhibitor aphidicolin [45] . These phenotypes were attributed to the functions of MCM8 and MCM9 in mitotic homologous recombination . To test if MCMDC2 had a defect in mitotic cell cycle due to a possible function in mitotic recombination we established MEFs from Mcmdc2-/- and Mcmdc2+/+ litter-mate embryos . MEFs of Mcmdc2-/- and wild-type mice did not differ significantly in their growth rate or in their sensitivity to aphidicolin ( Fig 2a and 2b ) . This suggests that unlike MCM8 and MCM9 , MCMDC2 does not play an important role during mitotic growth . Importantly , while we observed no obvious somatic defects , both sexes of Mcmdc2-/- mice were infertile ( no pups after 113 breeding weeks , n = 3 males and n = 3 females ) because both oogenesis and spermatogenesis were blocked ( Fig 2c–2e ) . Ovaries of 6 weeks old Mcmdc2-/- females were atrophic , barely discernible , and completely devoid of oocytes ( n = 3 mice , Fig 2c and 2e ) . This was due to an apparent loss of oocytes perinatally or soon after birth , as oocytes were still present in ovaries of fetal and newborn Mcmdc2-/- mice . Spermatogenesis takes place within testis tubules , which can be found at 12 distinct stages of the seminiferous epithelial cycle ( stages I-XII ) . Each stage is identified by the distinct combinations of spermatogenic cells found within [53] . We observed no late prophase spermatocytes and postmeiotic cells in Mcmdc2-/- males . This was due to apoptosis of spermatocytes in epithelial cycle stage IV testis tubules ( Fig 2d ) ; in the wild type , stage IV tubules contain spermatocytes at the mid pachytene stage . Consistent with a complete elimination of spermatocytes at stage IV no histone H1T ( a late prophase marker [54] ) positive cells were found in Mcmdc2-/- testis tubules ( n>200 tubules , Fig 2d ) . In contrast , histone H1T was increasingly expressed in wild-type spermatocytes beyond stage IV as expected . Elimination of meiocytes at the observed stages indicated a possible defect in meiotic recombination in Mcmdc2-/- mice . Specifically , persistent asynapsis and failure in DNA break repair is known to elicit elimination of spermatocytes and oocytes in stage IV testis tubules and perinatal ovaries , respectively [55–58] . Therefore we examined synaptonemal complex formation and markers of meiotic recombination in Mcmdc2-/- meiocytes . While chromosome axes readily formed ( as judged by SYCP3 staining ) , synaptonemal complex formation was severely defective in Mcmdc2-/- meiocytes ( as shown by disrupted SYCP1 localization along chromosome axes ) ( Fig 3 ) . In the most advanced stages full axes formed along the core of each chromosome in Mcmdc2-/- spermatocytes . Fully assembled chromosome axes can be observed from late zygotene to diplotene in wild-type spermatocytes . Given that Mcmdc2 -/- spermatocytes were eliminated at a stage equivalent to wild-type mid-pachytene we postulate that Mcmdc2-/- spermatocytes with fully formed axes reached a prophase stage equivalent to wild-type late zygotene to mid pachytene , hence we refer to this stage in Mcmdc2 -/- spermatocytes as late zygotene-pachytene . In the wild type , all late zygotene spermatocytes had partially synapsed chromosomes and by early pachytene all autosomes fully synapsed ( n = 101 cells , of which 20% were late zygotene and 80% were early pachytene , Fig 3b ) . In contrast , in Mcmdc2 -/- spermatocytes , synapsis of all chromosomes was never observed ( n>500 spermatocytes ) as chromosome axes were unaligned and remained mostly unsynapsed in late zygotene-pachytene . The transverse filament component SYCP1 , which marks synapsed axes [59] , was detectable in a punctate pattern at very low levels along unsynapsed chromosome axes , or was detected only at a few foci at intersections of chromosome axes in 36% of late zygotene-pachytene Mcmdc2 -/- spermatocyte nuclear spreads ( n = 412 cells , Fig 3b ) . In the rest of the spermatocytes , stretches of SYCP1 were detected along sections of juxtaposed axes , but the numbers of these synaptonemal complexes were generally low ( a median number of 4 SYCP1 stretches in each spermatocyte , n = 131 cells ) . Chromosomes with different axis lengths were engaged in synapsis with each other , and chromosomes formed synapsis with more than one partner indicating that synaptonemal complexes frequently formed between non-homologous chromosomes ( 53% of synaptonemal complex stretches were unequivocally identified as non-homologous , n = 232 synaptonemal complex stretches in 62 spermatocytes ) . Nevertheless , in a significant fraction of cells ( 14 out of 116 cells ) we observed apparently fully synapsed chromosomes ( median 1 fully synapsed chromosome , range 1–4 ) . We observed a similar chromosome alignment and synaptonemal complex formation defect in Mcmdc2-/- oocytes that were collected from 16 or 18dpc fetuses . At these stages most wild-type oocytes were in early ( 16dpc ) or late ( 18dpc ) pachytene stages , and all chromosomes in oocytes with fully formed axes were either fully synapsed ( 16dpc: 89% , n = 500 , 18dpc: 90% , n = 325 oocytes ) or partially synapsed ( 16dpc: 11% , 18dpc: 10% ) ( Fig 3d ) . In contrast , a large fraction of Mcmdc2 -/- oocytes with fully formed axes ( 16dpc: 55% , n = 136 , 18dpc: 41% , n = 174 oocytes ) lacked synapsis completely or formed only punctate/very short stretches of synaptonemal complexes at intersections of axes ( Fig 3d ) . Stretches of synaptonemal complexes formed in 45% ( 16dpc ) or 59% ( 18dpc ) of oocytes . Among those oocytes with SYCP1 stretches , the median number of stretches was 4 at 16dpc ( n = 61 ) and 5 at 18dpc ( n = 103 ) . We also observed apparently fully synapsed chromosomes in 10 out of 61 ( 16dpc ) or 73 out of 103 ( 18dpc ) oocytes where SYCP1 stretches were observed . In cells which had fully synapsed chromosomes , a median number of 1 fully synapsed chromosome was observed at 16dpc ( n = 10 oocytes ) and 2 at 18dpc ( n = 73 ) . The highest number of fully synapsed chromosomes we observed was 8 . We detected low levels of punctate SYCP1 signals along unsynapsed axes in Mcmdc2-/- meiocytes in both sexes ( male: Fig 3a , female: Fig 3c ) , although this SYCP1 staining pattern was more obvious in oocytes . This suggested that synaptonemal complex transverse filament assembly was initiated , although synaptonemal complex formation mostly failed in the absence of effective homologue alignment in Mcmdc2 -/- meiocytes . A similar weak association of SYCP1 with unsynapsed chromosome axes has been described in DNA strand invasion-defective Dmc1-/- and Hop2-/- meiocytes [60] . Thus , SYCP1 accumulation along unsynapsed chromosome axes may be a general phenomenon that can occur when synaptonemal complex formation is initiated but cannot be completed along unpaired chromosome axes . Taken together , these observations showed that MCMDC2 is critical for homologue alignment and synaptonemal complex formation in both sexes . The observed defects in homologue alignment and synaptonemal complex formation suggested that early stages of recombination may be defective in the absence of MCMDC2 . Single-stranded DNA ends that are produced after DSB formation are bound by recombinases RAD51 and DMC1 , which form foci along chromosome axes . These foci have been defined as early recombination nodules by electron-microscopy and are thought to represent recombinosome complexes [61–66] . DMC1 and RAD51 promote strand-invasion of DNA ends into homologues [9–11] . This leads to synaptonemal complex formation , and as DNA repair progresses , the early recombinosomes/recombination nodules lose DMC1 and RAD51 and progress to become transitional recombinosomes/nodules [61 , 62 , 64 , 65] . Hence , quantification of DMC1 and RAD51 foci is informative about the number of unrepaired DNA breaks involved in early stages of recombination in meiocytes . We found that foci of both RAD51 and DMC1 accumulate with similar kinetics in wild-type and Mcmdc2-/- spermatocytes in leptotene and early zygotene stages of prophase ( Fig 4c and 4d ) . Numbers of RAD51 and DMC1 foci dropped as wild-type spermatocytes progressed to early-mid pachytene . In contrast , high numbers of RAD51 and DMC1 foci persisted in late zygotene-pachytene Mcmdc2-/- spermatocytes . The high RAD51 and DMC1 foci numbers in Mcmdc2-/- spermatocytes required SPO11 ( Fig 4a and 4b ) . This observation is consistent with the idea that a delay in the repair of SPO11-generated programmed DSB breaks causes accumulation of RAD51 and DMC1 foci in Mcmdc2-/- spermatocytes . To further test if meiotic DSB repair is delayed in the absence of MCMDC2 we detected phospho-serine 139 histone H2AX ( ɣH2AX ) , which accumulates on chromatin in response to unrepaired DNA breaks and asynapsis in meiotic cells . ɣH2AX decorated chromatin in leptotene and zygotene stages but largely disappeared from autosomal chromatin due to progression of DSB repair and synapsis in wild-type cells . It remained associated only with the chromatin of the largely unsynapsed sex chromosomes , which form the sex body at the early pachytene stage ( Fig 4e ) . In stark contrast to wild-type spermatocytes , late zygotene-pachytene Mcmdc2-/- spermatocytes failed to form sex bodies and ɣH2AX persisted on autosomal chromatin ( n>200 cells ) . Persistent widespread ɣH2AX accumulation on chromatin was dependent on SPO11 , as Mcmdc2-/- Spo11-/- spermatocytes formed only more localized ɣH2AX-rich chromatin domains , so called pseudo-sex bodies , which are characteristic of mutants defective in programmed DSB formation ( Fig 4e ) [55 , 67] . We found a similar delay in meiotic recombination in Mcmdc2-/- oocytes ( Fig 5 ) . We observed low numbers of DMC1 foci ( median 29 . 5 foci , n = 66 cells ) or RAD51 foci ( median 21 foci , n = 62 cells ) in late pachytene wild-type oocytes at 18dpc ( Fig 5b and 5d ) . Correspondingly , the chromatin of pachytene oocytes was depleted of ɣH2AX ( n>100 oocytes , Fig 5e ) . In contrast , both foci of RAD51 ( median 162 . 5 foci , n = 66 cells ) and DMC1 ( median 173 foci , n = 64 cells ) persisted along unsynapsed axes and ɣH2AX accumulated to high levels throughout the genome ( n>100 oocytes , Fig 5e ) in oocytes from Mcmdc2-/- females . The combination of these observations suggests that MCMDC2 is not required for DSB formation or the loading of recombinases on single-stranded DNA ends . Yet , RAD51/DMC1 appears to be ineffective in promoting homologue alignment and DNA break repair is severely impaired in the absence of MCMDC2 . In wild-type meiosis , successful homologue alignment is accompanied by the appearance of axis-associated MSH4 foci that are thought to represent MSH4/5 ( MutSγ ) complexes within transitional recombinosomes/recombination nodules [38 , 61 , 64] . MutSγ is necessary for robust homologue pairing and alignment , most likely because it stabilizes strand invasion intermediates [23 , 36–39] . Hence , MSH4 foci are inferred to mark stabilized post-strand invasion recombination intermediates that are needed for efficient homologue alignment . Therefore , we tested if MSH4 forms axis-associated foci in Mcmdc2-/- meiocytes ( Fig 6a–6d ) . MSH4 foci numbers were significantly lower in Mcmdc2-/- spermatocytes ( median 11 . 5 foci , in late zygotene-pachytene cells , n = 48 ) and oocytes ( median two foci in 16dpc late zygotene-pachytene oocytes , n = 48 ) than in wild-type meiocytes ( median 80 foci in early-mid pachytene spermatocytes , n = 49 , median 139 foci in 16dpc late zygotene and pachytene oocytes , n = 64 ) . MSH4 foci numbers were marginally higher in late zygotene-pachytene than leptotene Mcmdc2-/- spermatocytes ( p = 0 . 0138 , Mann Whitney test ) . This may indicate that MSH4-marked intermediates still form with low efficiency in the absence of MCMDC2 . However , we observed an increasing punctate anti-MSH4 signal throughout the nuclei of wild-type meiocytes upon progression to pachytene ( see Fig 6a upper panel ) . This pan-nuclear signal is unlikely to represent MutSγ complexes bound to recombination intermediates . Thus , a “background” anti-MSH4 signal in Mcmdc2-/- spermatocytes may provide an explanation for the low MSH4 foci counts , which show a small increase upon progression from leptotene to late zygotene-pachytene . Regardless , the strongly reduced MSH4 foci numbers of Mcmdc2-/- meiocytes suggest severe impairment in MutSγ function and/or in a recombination step that precedes recruitment of MutSγ to recombination intermediates . Most MSH4-marked intermediates are repaired as non-crossovers in late pachytene , but a minority of them ( at least one per homologue pair and on average 23 per cell ) is thought to develop into MLH1-marked late recombinosomes ( defined as late recombination nodules by electron-microscopy ) [61 , 62 , 64 , 65 , 68] , which are sites of future crossovers [5 , 69 , 70] . Consistent with an impairment in MutSγ-containing recombinosome formation , and consistent with the elimination of spermatocytes in mid-pachytene , we found no MLH1 foci in Mcmdc2-/- spermatocytes ( n = 35 spermatocytes , Fig 6e ) . We also detected a similar defect in MLH1 foci formation in Mcmdc2-/- oocytes that had full axis ( n = 23 oocytes at 18dpc , and n = 47 oocytes at 20 . 5dpc/newborn , Fig 6f ) . Thus , the recombination defect in Mcmdc2-/- meiocytes ultimately prevents the formation of MLH1 foci , which likely represent precursors of a large majority of meiotic crossovers . While analyzing synaptonemal complexes in Mcmdc2-/- meiocytes , we noted that synaptonemal complex formation was more severely affected in Mcmdc2-/- meiocytes than in Spo11-/- meiocytes ( Fig 7 ) . Spo11-/- meiocytes lack programmed DNA breaks , and thus fail in homologue alignment and homologous synaptonemal complex formation . Nevertheless , synaptonemal complexes extensively formed between non-homologous chromosomes often creating a meshwork of interconnected chromosomes in Spo11-/- meiocytes of the most advanced stages ( Fig 7a middle panel ) . Our observations suggest that MCMDC2 is needed for meiotic DSB repair and progression beyond the early stages of recombination . Thus , early recombination intermediates may inhibit non-homologous synaptonemal complex formation in Mcmdc2-/- meiocytes . Alternatively , it is possible that MCMDC2 has a DSB-independent function that is needed for non-homologues synaptonemal complex formation when homologue alignment is defective . To distinguish between these possibilities we tested the epistatic relationship between Mcmdc2 and Spo11 . We reasoned that if non-homologous synapsis formation was similarly limited in Mcmdc2-/- and Spo11-/- Mcmdc2-/- meiocytes then this would indicate a DSB-independent role for MCMDC2 in non-homologous synaptonemal complex formation . Conversely , extensive non-homologous synaptonemal complex formation in Spo11-/- Mcmdc2-/- meiocytes would indicate a role for SPO11 and SPO11-dependent recombination intermediates in the inhibition of non-homologous synaptonemal complex formation in Mcmdc2-/- meiocytes . We found that Mcmdc2-/- did not significantly reduce non-homologous synaptonemal complex formation in a Spo11-/- background , and accordingly , we observed significantly more non-homologous synaptonemal complex stretches in Spo11-/- ( p = 0 . 0001 , Mann Whitney test ) and Spo11-/- Mcmdc2-/- ( p = 0 . 0001 , Mann Whitney test ) spermatocytes than in Mcmdc2-/- spermatocytes ( Fig 7b ) . Thus , MCMDC2 is not required for non-homologous synapsis that forms in the absence of SPO11 and programmed DSBs . We conclude that Spo11 is epistatic to Mcmdc2 in “erroneous” non-homologous synaptonemal complex formation , and that SPO11-dependent recombination intermediates , which fail to promote homologous synaptonemal complex formation in the absence of MCMDC2 , most likely interfere with the formation of extensive non-homologous synapsis in Mcmdc2-/- spermatocytes .
Our work revealed that the MCM domain-containing protein MCMDC2 is essential for meiotic recombination , and hence gametogenesis . In contrast to MCM8 and MCM9 , which have been implicated in homologous recombination during stalled replication fork restart and interstrand crosslink repair in mitotically growing cells [45–48] , we found no evidence of an important role for MCMDC2 in mitotic cells ( Fig 2a and 2b ) . We found that Mcmdc2-/- meiocytes were defective in recombination-mediated repair of programmed meiotic DSBs , alignment of homologues , and synaptonemal complex formation . DSB repair and synaptonemal complex formation are mutually dependent on each other in mammalian meiosis [24–28 , 37–39 , 60 , 71 , 72] . Thus , the severe synapsis formation defect observed could either be a cause or a consequence of the failed DNA DSB break repair in Mcmdc2-/- meiocytes . We favor the hypothesis that the primary role of MCMDC2 is in DSB repair and not in synaptonemal complex formation . In support of this hypothesis , while synaptonemal complex formation is not required for correct alignment of homologue axes [24–28] , MCMDC2 and initial steps of recombination that involve the formation of stable inter-homologue strand invasion intermediates are required [37–39 , 60 , 71 , 72] . This implicates MCMDC2 in early recombination steps that are needed for homologue pairing . Additional support is provided by the observation that DNA DSB repair seems to progress further in synaptonemal complex-defective mutants than in Mcmdc2-/- meiocytes . RAD51/DMC1-marked early recombinosomes seem to develop into MSH4-marked transitional recombinosomes in mutant spermatocytes that lack structural components of the synaptonemal complex [24–26] . MSH4 is thought to stabilize inter-homologue recombination intermediates [11 , 23 , 35 , 36] , which is likely important for the extensive homologue pairing that takes place in synaptonemal complex deficient meiocytes . In contrast , MSH4 foci counts remained low in Mcmdc2-/- meiocytes , indicating an earlier impairment in recombination that provides a likely reason for the observed failure in homologue alignment . Thus , defective synaptonemal complex formation cannot account for the observed defect in recombination in Mcmdc2-/- meiocytes . These observations strongly indicate that defective synaptonemal complex formation is the consequence of failed recombination in Mcmdc2-/- meiocytes , and not vice versa . Consistent with this conclusion , we found that MCMDC2 was not required for synaptonemal complex formation in the DSB formation defective Spo11-/- spermatocytes , which form extensive non-homologous synapsis . This observation suggests that MCMDC2 is not involved directly in synaptonemal complex formation , although we cannot formally exclude the possibility that MCMDC2 plays a direct role specifically in homologous synaptonemal complex formation in a DSB formation-proficient background . Synaptonemal complex formation was much more limited in Mcmdc2-/- than in the DSB formation defective Spo11-/- and Spo11-/- Mcmdc2-/- meiocytes . Interestingly , synaptonemal complex formation is also reduced in strand-invasion defective Dmc1-/- and Hop2-/- meiocytes as compared to the DSB formation defective Spo11-deficient meiocytes [60] . Thus , accumulation of SPO11-dependent recombination intermediates may interfere with “erroneous” non-homologous synaptonemal complex formation in mutants where homologue pairing is defective due to an early block in recombination . Recombination intermediates might also inhibit non-homologous synaptonemal complex formation in unperturbed meiosis , which could help to ensure that synaptonemal complexes form between homologous substrates . Unrepaired DSBs may inhibit non-homologous synaptonemal complex formation directly . Alternatively , unrepaired DSBs may have an indirect effect by altering cell cycle-progression . Although , spermatocytes are eliminated in stage IV testis tubules in both DSB repair defective ( e . g . Dmc1-/- , or Mcmdc2-/- ) and the DSB formation defective Spo11-/- spermatocytes , it has been proposed that Spo11-/- spermatocytes progress further in meiotic prophase [55 , 73] . This is because the mid-late pachytene marker histone H1T was observed in Spo11-/- spermatocytes but not in DSB repair defective spermatocytes , which indicates that unrepaired DSBs likely delay progression through meiotic prophase [55 , 73] . The accumulation of RAD51 and DMC1 foci in Mcmdc2-/- meiocytes indicates that DNA ends were resected , and single-stranded DNA overhangs recruited strand-invasion promoting recombinases at break sites . Yet , these inferred RAD51 and DMC1 coated single-stranded overhangs were unable to efficiently promote homologue pairing . This might indicate that RAD51 and DMC1 recombinases cannot promote strand invasions effectively in the absence of MCMDC2 . Alternatively , strand invasions and D-loop formation may still occur , but these recombination intermediates are not stabilized sufficiently to ensure alignment of homologues and the formation of extensive synaptonemal complexes . The observation that MSH4 foci numbers are low in Mcmdc2-/- meiocytes is consistent with both of these scenarios . In the former scenario , recombination intermediates that could recruit the MutSγ complex would not form in Mcmdc2-/- meiocytes , hence MSH4 foci could not form either . In the latter scenario , accumulation of MutSγ complex at strand invasion intermediates/D-loops would be defective in Mcmdc2-/- meiocytes . Hence , these recombination intermediates would be unstable and might be dissolved by helicases , e . g . the BLM helicase , which has been proposed to antagonize MutSγ in its function of stabilizing recombination intermediates [31 , 35] . It follows , that MCMDC2 would play an important role in MutSγ function in the latter scenario . MCM proteins are AAA+ ATPases that form hexameric rings on duplex DNA and promote the melting of double-stranded DNA in an ATP dependent-manner ( reviewed in [40] ) . MCM2-7 are primarily involved in the initiation of DNA replication [74] , but MCM8 and MCM9 are particularly important for homologous recombination in mitotic cells in vertebrates [45–48] . MCM8 and MCM9 interact , and are thought to also form hexameric helicase complexes [45 , 46 , 48] . MCM8 and MCM9 promote DNA repair by facilitating resection of DNA ends at DSB sites [47] , promoting an as yet undefined post-strand-invasion steps of recombination [45 , 46] , and melting DNA at sites of mismatches during mismatch repair in somatic cells [75] . Despite being crucial for recombination and mismatch repair in mitotically dividing cells , MCM9 does not have an essential role in meiotic recombination [45 , 76] . Thus , MCM8 functions independent of MCM9 in meiosis . The Drosophila orthologues of MCM8 and MCMDC2 form a protein complex that is presumed to stabilize strand invasion intermediates of recombination specifically in meiosis [43] . Furthermore , the meiotic phenotypes of Mcm8-/- [45] and Mcmdc2-/- mice appear very similar , although no data was reported on MutSγ behavior in Mcm8-/- . Thus , it is tempting to speculate that MCM8 and MCMDC2 collaborate in mouse meiosis , and that MCMDC2 replaces MCM9 in MCM8-containing helicase complexes in meiosis . Curiously , the sequences of Walker A and B motifs , which are domains that are necessary for the ATPase activity of MCMs , are apparently not conserved in either Drosophila or mammalian MCMDC2 proteins [43] . This suggests that MCMDC2 is unlikely to function as an ATPase , but it may function as a modulator in putative meiotic helicase complexes that contain other MCMs ( e . g . MCM8 ) with an active ATPase domain . Interestingly , the Drosophila MCM8 orthologue , REC , was proposed to facilitate repair DNA synthesis during meiotic recombination , because meiotic gene conversion tracks were significantly shortened in rec mutants [41] . Putative mammalian MCM8/MCMDC2-containing helicase complexes may have similar functions . One possibility could be that MCMDC2 promotes unwinding of the invaded DNA at sites of strand invasions . This could facilitate the formation of extended strand invasions , and/or may be needed for efficient DNA repair-synthesis starting from the 3´end of the invading strands . These hypothesized functions would be expected to stabilize strand invasions . Unwinding invaded homologous sequences to promote the formation of extended and stable D-loops might be particularly important for inter-homologue recombination during meiosis . The reason is that mismatches that can occur between homologues would likely interfere with extension of D-loops thereby antagonizing the stabilization of inter-homologue strand-invasion intermediates . The observation that MCMDC2 is required for the accumulation of MSH4 at recombination intermediates during meiosis raises the interesting possibility that a putative MCMDC2-containing helicase complex and the MutSγ complex may physically interact and collaborate in stabilizing D-loops . At sites of DNA mismatches , MCM9 forms a complex with MSH2 and MSH6 , which are homologs of MutSγ components MSH4 and MSH5 , and the complex of these proteins is thought to be crucial for correct mismatch repair in mitotically active cells [75] . Thus , it is possible that functional and/or physical interaction between MSH proteins and “DNA repair-promoting” MCMs is a conserved principle in distinct DNA repair pathways . Relevant to this point is the observation that the MutSγ complex is missing from Schisophora , a taxon that includes Drosophila [43] . It has been proposed that a complex of REC ( Drosophila melanogaster MCM8 ) and MEI-217/218 ( two Drosophila melanogaster orthologues of MCMDC2 ) proteins substitutes for the functions of the missing MutSγ complex in antagonizing BLM helicase , stabilizing strand invasion intermediates , and promoting crossover formation in meiosis in Drosophila [43] . The REC/MEI-217/218 complex may have been capable of replacing MutSγ in Schisophora because MCMDC2-containing complexes and MutSγ might have interacted and had shared functions in stabilizing DNA strand invasion intermediates in ancestral taxa where both of these complexes existed . Thus , loss of MutSγ in Schisophora may have required only a modification to an already pre-existing ( and possibly conserved ) function in MCMDC2-containing complexes . This speculative scenario would be certainly consistent with a putative conserved functional interplay of MutSγ and MCMDC2 in stabilizing recombination intermediates during mammalian meiosis . Thus , an important aim of future studies of MCMDC2 functions will be to address if MCMDC2 forms helicase complexes with other MCMs and if these complexes interact physically and functionally with MutSγ to stabilize D-loops .
To test Mcmdc2 expression in testes of wild-type and Mcmdc2-/- mice , RNA was isolated and RT-PCR was performed as described earlier [20 , 77] . The RNA of the somatic tissue mix in ( Fig 1a ) originated from 17 distinct tissues: liver , brain , thymus , heart , lung , spleen , kidney , mammary gland , pancreas , placenta , salivary gland , skeletal muscle , skin , small intestine , spinal cord , tongue and uterus . The sequence of transcript-specific primers for RT-PCR were: Mcmdc2 1R ( Fig 1a ) 5’-CGTTCCCTGTTGCAGTCTCT Mcmdc2 1F ( Fig 1a ) 5’-CCCCACACAGCAAAAGTTCC s9for ( Fig 1a ) 5’-GGCCAAATCTATTCACCATGC s9rev ( Fig 1a ) 5’-TAATCCTCTTCCTCATCATCAC Mcmdc2 Exon3 fw ( Fig 1d , exon 3/4 ) 5’-ATTCAAAGCAGAGTTATGCTG Mcmdc2 Exon4 rv ( Fig 1d , exon 3/4 ) 5’-TTGAGTTTCAGTCTGTAACTGT Mcmdc2 Exon5 fw ( Fig 1d , exon 5/6 ) 5’-ATCAATATTGTGCTGAAGTTAAC Mcmdc2 Exon6 rv ( Fig 1d , exon 5/6 ) 5’-ACCAAGTACTCTAAATTTTCTGT Mcmdc2 Exon6 fw ( Fig 1d , exon 6/7 ) 5’-GATTTCAGTATGTGAGAGTCC Mcmdc2 Exon7 rv ( Fig 1d , exon 6/7 ) 5’-CTCTTAGGAAAATACCAAGTGA Mcmdc2 Exon8 fw ( Fig 1d , exon 8/9 ) 5’-ATGAACTAGTGAATAAGATGAAAA Mcmdc2 Exon9 rv ( Fig 1d , exon 8/9 ) 5’-CTGTCTACAAGCAGAGTGTC Mcmdc2 Exon10 fw ( Fig 1d , exon 10/11 ) 5’-ACTTTTGAATTTTAGCATGAATCT Mcmdc2 Exon11 rv ( Fig 1d , exon 10/11 ) 5’-CATCTGACCAATCAGAGTACT Mcmdc2 was targeted in JM8A3 . N1 . C2 embryonic stem ( ES ) cells by the EUCOMM-IKMC project ( project: 93238 , ES line:HEPD0781_2_C06 and project: 118859 , ES line: HEPD0800_2_F07 ) . Targeting was based on a so called ‘knockout first’ multipurpose allele strategy [78] ( Fig 1 ) . Chimeras were generated by laser assisted C57BL/6 morula injections with ES cell clones heterozygote for the Mcmdc2 insertion allele ( Fig 1c ) . Progeny of the chimeric animals were crossed to the outbred wild-type CD-1 mouse line , and to pCAGGs-FLPo [79] and PGK-Cre [80] transgenic mice to generate Mcmdc2 restored , Mcmdc2 deletion , and Mcmdc2 insertion–deletion alleles from the Mcmdc2 insertion allele ( Fig 1b ) . Mice were maintained on the outbred ICR ( CD-1 ) background . Mice were genotyped by PCR using tail-tip genomic DNAs . Genotyping primers: LacZfor 5’-TGGCTTTCGCTACCTGGAGAGAC LacZrev 5’-AATCACCGCCGTAAGCCGACCAC CreFw 5’-GCCTGCATTACCGGTCGATGCAACGA CreRv 5’-GTGGCAGATGGCGCGGCAACACCATT FlpOFw 5’-GCTATCGAATTCCACCATGGCTCCTAAGAAGAA FlpORv 5’-CAATGCGATGAATTCTCAGATCCGCCTGTTGATGTA o566 5’-GCAAGAAAACTATCCCGACC o1046 5’-CACAGTGAGGCCCAATATAAA o1047 5’-TCCACAGGAAAAGGCAAACG o1049 5’-GGTGCTAGCCCCTTCCTTTT o1050 5’-TCACTTGGTATTTTCCTAAGAG o1147 5’-TGAAAGTTGATATGAAACTGTATA o1163 5’-AAGGTTGTAGAATTACAGCAGC PCR product sizes: with LacZfor/LacZrev primers , Mcmdc2insertion and Mcmdc2insertion–deletion allele 208 bp , other alleles-no specific product; with o1046/o1047 primers , wild-type allele 244 bp , Mcmdc2insertion allele 225bp , Mcmdc2restored allele 225 bp , Mcmdc2deletion allele no product , Mcmdc2insertion–deletion allele no product; with o566/o1047/o1050 primers , wild-type allele 561bp , Mcmdc2insertion allele 542bp , Mcmdc2restored allele 542bp , Mcmdc2deletion allele no product , Mcmdc2insertion–deletion allele 751bp; with o1049/o1147 primers , wild-type allele 310bp , Mcmdc2insertion allele effectively not amplifiable product ( 7370 bp ) , Mcmdc2restored allele 467bp , Mcmdc2deletion allele no product , Mcmdc2insertion–deletion allele no product; with o1049/o1163/o1147 primers , wild-type allele 310 bp ( and 3421bp ) , Mcmdc2insertion allele effectively no amplifiable products ( 7370bp and 10462bp ) , Mcmdc2restored allele 467bp ( and 3559bp ) , Mcmdc2deletion allele 690 bp , Mcmdc2insertion–deletion allele effectively not amplifiable product ( 5683bp ) . FlpOFw/FlpORv were used to detect FlpO recombinase transgene ( 1500 bp ) , CreFw/CreRv were used to detect Cre recombinase transgene ( 750 bp ) . Mice carrying Spo11-null alleles were described earlier [6 , 7] . Histology in testis , analysis of RAD51/DMC1 foci or the synaptonemal complex were carried out in mice lines derived from both independent clones . The phenotypes of all the listed alleles were examined . No obvious differences were detected in testis histology , RAD51/DMC1 foci accumulation and synaptonemal complex formation between mice derived from the different ES clones and between the Mcmdc2 insertion/insertion , Mcmdc2 deletion/deletion , and Mcmdc2 insertion–deletion/insertion–deletion strains . We chose the HEPD0800_2_F07 derived Mcmdc2 insertion–deletion allele for complete phenotypic analysis , hence this line was used in all the reported experiments . Given that Mcmdc2 insertion–deletion/insertion–deletion mice lack three exons , which causes a frameshift we refer to this genotype as Mcmdc2 -/- . Mcmdc2 restored/ restored mice were fertile and their spermatocytes were indistinguishable from wild-type spermatocytes reconfirming the specificity of the observed phenotypes in the Mcmdc2 insertion/insertion , Mcmdc2 deletion/deletion , and Mcmdc2 insertion–deletion/insertion–deletion strains . Whenever possible , experimental animals were compared with littermate controls or with age-matched non-littermate controls from the same colony . All animals were used and maintained in accordance with the German Animal Welfare legislation ( “Tierschutzgesetz” ) , the Directive 2010/63/EU of the European Parliament and of the Council on the protection of animals used for scientific purposes and its German implementation ( Tierschutz-Versuchstierverordnung–TierSchVersV ) . All procedures pertaining to animal experiments were approved by the Governmental IACUC ( "Landesdirektion Sachsen” ) and overseen by the animal ethics committee of the Technische Universität Dresden . The license numbers concerned by the present experiments are DD24-5131/287/1 and 24–9168 . 24-1/2006-13 ( tissue collection without prior in vivo experimentation ) . Wild-type and Mcmdc2-/- mouse embryonic fibroblasts were derived from 12 . 5–14 . 5dpc embryos using standard procedures [81] . We plated 10 , 000 mouse embryonic fibroblasts in triplicates in 1ml DMEM ( GIBCO ) in 24-well plates . Live cells were counted using Miltenyi Biotec MACSQuant on day 3 , 6 , 9 , 13 and 15 of cultures without aphidicolin . For aphidicolin treatment , 10000 cells were plated and incubated with media containing 1μM aphidicolin for the first 24 hour of the culture . Live cells were counted on day 5 , 9 , 13 , 17 after plating . The media was changed every 3rd day for both types of cultures . In addition to antibodies that were previously described [20 , 21] we used two commercial antibodies: mouse anti-MLH1 ( IF 1:50 , BD Biosciences , order number 551092 ) and rabbit anti-MLH1 ( IF 1:50 , Calbiochem , order number D00122409 ) . We also used a chicken anti-SYCP3 antibody that was raised against a His-tagged version of a 99 amino acid-long ( from 13E to 111E amino acids ) peptide of SYCP3 , which we overexpressed in Escherichia coli and purified using metal ion affinity chromatography . IgYs from the yolk of eggs of immunized chicken were extracted using a published protocol [82] . Anti-SYCP3 IgYs were affinity purified on immunizing-antigen coupled NHS-Activated Sepharose 4 Fast Flow beads ( Cat#17-0906-01 , Amersham , GE Healthcare ) according to standard methods [83] . Preparation and immunostaining of testis-ovary cryosections and nuclear surface spreads of meiocytes were carried out as described before [20 , 21 , 84 , 85] . Recombination foci and synaptonemal complex stretches were counted manually on matched exposure images with the use of the count tool of Photoshop CS5 . We counted anti-RAD51 , -DMC1 , -MSH4 or -SYCP1 signals that were associated with SYCP3-marked chromosome axes , to avoid counting signals that do not represent genuine recombination foci ( RAD51 , DMC1 and MSH4 ) or synaptonemal complexes ( SYCP1 ) . Statistical analysis was carried out with GraphPad Prism 5 . For the comparison of independent samples , the two-tailed non-parametric Mann_Whitney ( two-sample Wilcoxon rank-sum ) test was used . | Each chromosome is present in two distinct but homologous copies in diploid organisms . To generate haploid gametes suitable for fertilization , these homologous chromosomes must segregate during meiosis . To ensure correct chromosome segregation , homologous chromosomes must align and become connected by inter-homologue crossovers during early meiosis in most taxa including mammals . Defects in these processes result in infertility and aneuploidies in gametes . Alignment of homologous chromosomes and crossover formation entail generation of DNA double-strand breaks and repair of DNA breaks by meiotic recombination . As part of the repair process , single-stranded DNA ends resulting from DNA breaks invade homologous DNA sequences and use them as repair templates . DNA strand invasion events lead to the alignment of homologous chromosomes , and serve as precursors for crossovers . We discovered that meiotic recombination critically depends on the helicase-related minichromosome maintenance domain containing 2 protein ( MCMDC2 ) . MCMDC2 likely promotes the formation and/or stabilization of DNA strand invasion events that connect homologous chromosomes . Thus , MCMDC2 is required for DNA breaks to effectively promote alignment of homologous chromosomes . This work reveals a crucial role for MCMDC2 in recombination in mammals , and constitutes an important step in understanding how recombination establishes connections between homologous chromosomes during meiosis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"meiosis",
"homologous",
"chromosomes",
"spermatocytes",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"germ",
"cells",
"oocytes",
"dna",
"sperm",
"homologous",
"recombination",
"animal",
"cells",
"chromosome",
"biology",
"proteins",
"recombinant",
"proteins",
"biochemistry",
"cell",
"biology",
"nucleic",
"acids",
"ova",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"dna",
"repair",
"dna",
"recombination",
"chromosomes"
]
| 2016 | Alignment of Homologous Chromosomes and Effective Repair of Programmed DNA Double-Strand Breaks during Mouse Meiosis Require the Minichromosome Maintenance Domain Containing 2 (MCMDC2) Protein |
We have investigated the pathogenicity of tsetse ( Glossina pallidipes ) -transmitted cloned strains of Trypanosoma brucei rhodesiense in vervet monkeys . Tsetse flies were confirmed to have mature trypanosome infections by xenodiagnosis , after which nine monkeys were infected via the bite of a single infected fly . Chancres developed in five of the nine ( 55 . 6% ) monkeys within 4 to 8 days post infection ( dpi ) . All nine individuals were successfully infected , with a median pre-patent period of 4 ( range = 4–10 ) days , indicating that trypanosomes migrated from the site of fly bite to the systemic circulation rapidly and independently of the development of the chancre . The time lag to detection of parasites in cerebrospinal fluid ( CSF ) was a median 16 ( range = 8–40 ) days , marking the onset of central nervous system ( CNS , late ) stage disease . Subsequently , CSF white cell numbers increased above the pre-infection median count of 2 ( range = 0–9 ) cells/µl , with a positive linear association between their numbers and that of CSF trypanosomes . Haematological changes showed that the monkeys experienced an early microcytic-hypochromic anaemia and severe progressive thrombocytopaenia . Despite a 3-fold increase in granulocyte numbers by 4 dpi , leucopaenia occurred early ( 8 dpi ) in the monkey infection , determined mainly by reductions in lymphocyte numbers . Terminally , leucocytosis was observed in three of nine ( 33% ) individuals . The duration of infection was a median of 68 ( range = 22–120 ) days . Strain and individual differences were observed in the severity of the clinical and clinical pathology findings , with two strains ( KETRI 3741 and 3801 ) producing a more acute disease than the other two ( KETRI 3804 and 3928 ) . The study shows that the fly-transmitted model accurately mimics the human disease and is therefore a suitable gateway to understanding human African trypanosomiasis ( HAT; sleeping sickness ) .
In human African trypanosomiasis ( HAT ) , the use of animal models has contributed enormously to what is currently known about the relationships between disease duration , parasite invasion of different body systems and the potential of resultant host clinical and biological changes as diagnostic and disease staging markers . Several host-parasite model systems have been developed , based on infection of various hosts with the livestock pathogen Trypanosoma brucei brucei and to a lesser extent the human pathogens T . b . rhodesiense and T . b . gambiense . Characterisation of these HAT models shows that the disease occurs in two stages irrespective of host: an early haemo-lymphatic trypanosome proliferation , and a late central nervous system ( CNS ) infection , indicating that the basic pattern is similar to the disease in humans . This is evidenced by demonstration of trypanosomes , first in the haemo-lymphatic system and later in the CNS of the mouse model with subsequent cerebral pathology [1] , [2] . Models based on larger mammals such as the chimpanzee T . b . rhodesiense model [3] , the vervet monkey T . b . rhodesiense model [4] and the sheep T . b . brucei model [5] , also follow a similar two-stage disease pattern . These , unlike rodents , allow collection of cerebrospinal fluid ( CSF ) that has been used to demonstrate elevation of white cell counts and total protein levels as indicators of CNS stage disease [6] . The KETRI vervet monkey model has been reported to closely mimic HAT clinically , immunologically and pathologically [4] , [7]–[9] . However , these previous studies were limited in scope in three important ways . Firstly , infections were initiated by intravenous inoculation ( syringe ) of bloodstream form trypanosomes as opposed to the natural human disease , which begins via the bite of a tsetse fly , with the intra-dermal inoculation of metacyclic trypanosomes . The difference between the two routes of infection has the potential to affect trypanosome virulence and subsequent disease pathogenesis that has been little explored to date . Second , disease progression has been monitored mainly in terms of clinical symptoms , gross pathology , histo-pathology and antibody responses [4] , [7] , with little reference to the development of blood pathology . Third , only a single strain of trypanosomes , KETRI 2537 [9] , has been adequately characterised even though trypanosome strains vary in the severity of pathogenesis and virulence [10]–[11] . The present study was designed to address these limitations and thus improve further the potential utility of the model and our understanding of pathogenesis in trypanosome infections . We characterised the pathogenicity of T . b . rhodesiense in vervet monkeys , following infection from the bite of a single tsetse fly , hence mimicking the natural route of infection in man . This allowed us to measure a range of parameters in blood and cerebrospinal fluid ( CSF ) including several that had not previously been studied . It was thus possible to measure the development of clinical complications of HAT infections , such as anaemia , more precisely . We describe the development of clinical pathology resulting from infection with four cloned strains of T . b . rhodesiense .
This study was undertaken in adherence to experimental guidelines and procedures approved by the Institutional Animal Care and Use Committee ( IACUC ) , the ethical review committee for the use of laboratory animals . Trypanosome isolates that were used in this study ( Table 1 ) were all initially obtained through collection of infected blood from patients in the western Kenya/eastern Uganda focus of endemic T . b . rhodesiense sleeping sickness ( historically known as the Busoga focus ) . All the isolates are maintained as cryo-preserved stabilates in the KARI-TRC ( formerly KETRI ) trypanosome bank . The isolates were included in the study on the basis of the year of isolation , to give a wide temporal distribution and the locality of isolation to give a wide spatial distribution within this sleeping sickness focus . The selected stabilates were cloned using the hanging drop method described by Herbert and Lumsden [12] . Male teneral tsetse flies ( Glossina pallidipes ) were obtained from the KETRI colony initially established with pupae from the Lambwe Valley of Kenya , which is part of the western Kenya/eastern Uganda focus of HAT . In order to initiate infection of tsetse flies with each trypanosome clone , four sub-lethally irradiated ( 600 rads , 5 minutes ) donor Swiss White mice were each inoculated intraperitoneally with 0 . 2 millilitres of the thawed T . b . rhodesiense stabilates , diluted in phosphate saline glucose ( PSG ) . At peak parasitaemia , typically approximately 108 trypanosomes per millilitre , a batch of 50 teneral flies were allowed to feed essentially as described [13] , and maintained thereafter on clean bovine blood by feeding via a silicon membrane . Thirty days after the infective blood meal , all the flies were chilled briefly and separated into individual fly cages . The flies with mature trypanosome infections were then identified by xenodiagnosis using Swiss White mouse . We were repeatedly unable to find trypanosomes in salivary probes on warm microscope slides [14] . Nine vervet monkeys ( Chlorocebus aethiops , African Green Monkeys ) of both sexes weighing between 2 . 7 and 5 . 2 kg were acquired from the Institute of Primate Research ( IPR ) in Kenya . They were housed in quarantine for a minimum of 90 days while being screened for evidence of disease , including zoonoses as described by Ndung'u and colleagues [8] . They were also dewormed and treated for any ectoparasite infestations . During the quarantine period , the animals became accustomed to staying in individual squeeze-back stainless steel cages and human handling . During quarantine and also while in the experimental animal wards , the monkeys were maintained on green maize , fresh vegetables ( bananas , tomatoes and carrots ) and commercial monkey cubes ( Monkey cubes , Unga Ltd , Kenya ) , fed twice daily ( 9 . 00–9 . 30 am and 3 . 00–3 . 30 pm ) , and given water ad libitum . After the expiry of the 90 days quarantine , the study animals were then transferred to experimental wards and acclimatised for a further two weeks prior to commencement of pre-infection data collection . The monkeys were randomly allocated into four experimental groups , each containing at least one male and one female , for infection with T . b . rhodesiense clones as follows: KETRI 3741 ( three monkeys , #s . 476 , 515 , and 536 ) , KETRI 3801 ( monkey #s . 523 , 579 ) , KETRI 3804 ( monkey #s 556 and 574 ) and KETRI 3928 ( monkeys #s 554 and 555 ) . Pre-infection ( baseline ) data was collected over a period of 14 days after which each monkey was infected by allowing one tsetse fly , confirmed trypanosome positive through mouse infectivity tests , to feed on a shaved part of its thigh , while the monkey was under ketamine Hcl ( Rotexmedica , Trittau Germany ) anaesthesia . Before and following the infective tsetse bite , the monkeys were monitored for activity , posture , demeanour and general clinical presentation on a daily basis . Appetite was assessed daily , by scoring the proportion of the daily feed ration consumed by each monkey on a scale of 0 ( no food eaten ) , 1/4 , 1/2 , 3/4 and 1 ( full ration eaten ) . Parasitaemia was assessed daily using the method of Herbert and Lumsden [12] , using heparinised capillary blood drawn from the ear vein , starting from the third day after infection . Every four days , the monkeys were sedated using ketamine hydrochloride ( 10–15 mg per kg body weight intramuscularly ) after which a detailed clinical examination was carried out and 2 ml of venous blood ( femoral ) sampled for a full haemogram . Every eight days , a CSF sample was also collected through lumbar puncture for assessment of CNS parasitosis and white cell numbers . The experiment was terminated through humane euthanasia at extremis . An individual animal was judged to be in extremis when for three consecutive days it was either unable or reluctant to perch , had very low feed intake ( <1/4 of daily ration ) , and in addition had signs of advanced late stage disease ( e . g somnolence ) . Euthanasia was carried out using 20% pentobarbitone sodium ( Euthatal , Rhone Merieux ) . Cerebrospinal fluid white cell counts ( WCC ) and total trypanosome numbers were concurrently counted using a Neubert chamber as previously described [6] , [8] . Immediately after every sampling session ( not exceeding one hour ) , total red blood cell ( RBC ) and related indices , white cell numbers and differential , platelet ( thrombocyte ) counts and associated parameters were determined using an AC3diff T Coulter counter ( Miami , Florida , USA ) . Data was entered and managed using Microsoft Excel ( Version 2003 ) . Statistical analysis was conducted using Statview for Windows Version 5 . 0 . 1 ( SAS Institute Inc , 1995–1998 , Cary , NC ) . The behaviour of the four trypanosome strains was analysed and is presented as tables and or graphs representing time bound changes in individual infected monkeys' clinical , haematological and cerebrospinal fluid pathology data . In addition , descriptive statistics [mean ( and the corresponding 95% confidence intervals , CI ) , or medians and range] were derived for the entire group of nine monkeys . In addition to derivation of descriptive data , haematology data was further analysed using repeated measures ANOVA . Finally , Spearman's correlation coefficients were determined to assess the strength of association between CSF trypanosome and white cell numbers .
In this study , infection of vervet monkeys was initiated by the bite of a single infected tsetse fly . To our knowledge , this represents the first time single fly transmission of T . b . rhodesiense clones in vervet monkeys ( or any other primate model ) has been achieved , hence establishing a model that more accurately mimics the transmission of sleeping sickness as it occurs in humans . Information on natural HAT relies on data provided in case reports [15]–[16] and sometimes re-analysis of retrospective clinical , epidemiological and pathology data [17]–[18] . Such data are naturally limited on the questions of disease onset and duration as there are only a small number of case reports in which the patient could accurately remember the exact time of being bitten by a tsetse fly [16] , [18] . In any case this is not readily accessible information for inhabitants of endemic areas , where tsetse fly challenge is continuous . This study has allowed us to generate information on the pathogenesis of HAT in a manner that more accurately mimics human disease , and facilitates documentation of data from precise sampling points during the course of a tsetse transmitted infection . All four T . b . rhodesiense clones that were selected for this study were successfully transmitted through tsetse flies in the initial step of the model development protocol , a result that is consistent with the very good vectorial capacity of G . pallidipes for both human and animal trypanosomiasis in eastern Africa . Tsetse flies carrying mature infections were identified by xenodiagnosis using Swiss White mice . The pre-patent period in these mice ( data not shown ) , and subsequently in monkeys ( Table 2 ) , showed considerable variation from host to host , consistent with observations in HAT patients [18]–[19] . In contrast , mice and vervet monkeys that are infected with T . b . rhodesiense via syringe passage show less variation in pre-patent period [4] , presumably because the inoculum in these is usually well-defined , while tsetse flies are estimated to inject 0–40 , 000 ( mean , 3 , 200 ) infective metacyclics [20] . Trypanosomes were detected in 7 of the monkeys within 5 days after infection ( Table 2 ) , indicating that the movement of trypanosomes from the site of the fly bite to the systemic circulation occurred quickly . This is remarkable , because it must be associated with the transition from non-proliferating metacyclics to rapidly dividing long slender bloodstream forms , clearly a survival strategy for the parasites . This movement was independent of the development of chancre , which was only observed in 5 monkeys . Chancres have been observed in HAT patients in whom these swellings are estimated to occur within 5–15 days of an infective fly bite [20] , consistent with our data ( Table 2 ) . During the formation of the chancre , the metacyclics transform to rapidly dividing bloodstream form trypanosomes while the tissue at the inoculation site mounts a reaction characterized by a marked infiltration with polymorphonuclear leucocytes [21]–[22] . The immune reaction generated at the chancre is responsible for development of specific immunity against the variable antigen type of metacyclics [23] . The finding that the severity of the clinical disease differed between individual monkeys that were infected with the same strain emphasized the likely role of host immunity on disease outcome . The ability/inability of the host to control parasitaemia and its effect on disease duration is further indicated by the observation that parasitaemia patterns showed more fluctuation ( clearly marked waves ) in individuals with longer disease duration than in those with shorter durations ( Figure 1 ) . A similar trend was observed in mice ( data not shown ) , consistent with previous reports [11] . In our study , clone 3801 produced a more acute disease , while clone 3928 manifested the most chronic disease; the other two clones were intermediate . These observations suggest that the parasites have intrinsic properties that , in part , determine virulence . Host factors also contribute to the disease profile , as evidence from variations in animals infected with the same clone indicates . A study of a number of isolates from eastern Uganda by Smith and Bailey [24] in mice showed that distinct acute and chronic strains of T . b . rhodesiense circulate in the focus and each strain is related to a given zymodeme . However , apart from individual parasite variations there were no features that could distinguish the Ugandan from Kenyan isolates . This supports the view that the four strains used in this study belong to the same endemic focus characterized by pockets of specific zymodemes with distinct clinical manifestations [24] . Similar diversities in clinical manifestations have been observed in HAT patients infected with T . b . rhodesiense [25] and T . b . gambiense [26] showing that animal models accurately mirror the situation in humans Haematology results showed that anaemia developed early in the monkey infections; the decline in relevant parameters was detected as early as 8 dpi . However , the rate of decline of RBC and associated parameters was much slower than in T . brucei infected mice in which the numbers of circulating erythrocytes can fall by up to 50% within a week after infection [27] . Anaemia is a common occurrence in both T . b . rhodesiense and T . b . gambiense forms of sleeping sickness [15]–[16] , [26] , [28] , similar to the case in vervet monkeys . However , determination of the rates of decline of RBC and associated parameters , is not possible in humans since neither the date of infection nor the pre-infection values in individual patients are known . The type of anaemia reported in our study , microcytic hypochromic , was different from the normocytic anaemia observed in T . brucei infected mice [29] or Nigerian mongrel dogs [30] during the acute phase of T . b . brucei infection . Microcytic hypochromic anaemia has previously been associated with iron deficiency [31] and could perhaps be related to failure of iron incorporation into red cell precursors or inefficient recovery of iron from the phagocytosed RBC , features which are common during acute trypanosomiasis [32] . Determination of the type of anaemia found in infected humans is complicated by presence of concurrent infectious and nutritional conditions [29] . This is compounded by the lack of appropriate haematology analysers in endemic areas , and has therefore not been systematically determined to our knowledge . The severe progressive thrombocytopaenia reported in our study mirrors that found in other T . b . rhodesiense animal models [30] and human cases of sleeping sickness [28] . These findings indicate that unlike in mild cases of iron deficiency anaemia that are accompanied by thrombocytosis , the anaemia of trypanosomiasis in both humans and animals is severe and could be related to a deficit in the production of thrombopoetin [33] . Similarly , leukocyte changes are broadly consistent with findings from other non–human primate studies [3]–[4] and humans [17] . However , the strong granulocyte response that coincided with the day of first detection of trypanosomes in peripheral blood ( median = 4 dpi ) has not been reported before , perhaps due to the lower frequency of sampling employed in other studies . Importantly , the presence of multiple peaks of white cells during the course of the infection suggests that in spite of the widely reported immunosuppressive effects of trypanosome infections , myeloid precursor cells retain the ability to proliferate in response to dominant parasite VSG's expressed during the course of the disease . This is in agreement with findings that some bone marrow stem cells survive the damage caused by trypanosomes and retain the ability to repopulate the animal [34] and may account for the observation of very late stage leucocytosis in some individuals but not others . The first evidence of trypanosomes in the CSF was on day 16 ( range 8–40 ) days ( Table 2 ) . This event is recognised by WHO [6] as a definitive marker for the onset of late stage infection . The timing of CSF parasitosis was largely similar to earlier observations in the syringe passage monkey infections where the blood-brain-barrier ( BBB ) was breached within 7–21 days [8] , [35] . Clone 3928 produced the most chronic infection of all isolates and , in monkey 554 , was only detected in the CSF on day 40 after infection . Some T . b . rhodesiense isolates from south-eastern Africa foci and some from eastern Uganda have been reported to cause a chronic HAT infection in humans , taking relatively long to invade the CNS [24] , [36] . One of the recognized markers of CNS pathology is the presence of raised numbers of leucocytes in the CSF above the background ( pre-infection ) levels [6] , [8] , 37–38 . Indeed , there was positive linear association between trypanosomes in the CSF and white cell changes , suggesting that both events are primarily determined by a single cause , possibly damage to the blood-brain barrier . The numbers of trypanosomes in CSF increased dramatically as disease progressed , and clinical symptoms of disease necessitated individuals to be removed from the study on ethical grounds , marking the terminal stage . The results of this study establish a cyclic T . b . rhodesiense model that more closely resembles the East African form of HAT . Although T . b . gambiense causes a more insidious slowly developing disease , the essential features including fever , loss of appetite , headache , fatigue , weight loss , leg paresthesis , gait difficulties and daytime somnolence are similar to symptoms observed in patients infected with T . b . rhodesiense [18] , [39] . Thus , this disease model in which the infection is induced using the bite of a single fly can better represent the complex pathogenesis of natural HAT . This model allows more precise timing of events , such as date of infection , and the clinical and haematology features that follow . Consequently , it is hoped that the new model will gain application and facilitate studies that require good precision . | Sleeping sickness is caused by a species of trypanosome blood parasite that is transmitted by tsetse flies . To understand better how infection with this parasite leads to disease , we provide here the most detailed description yet of the course of infection and disease onset in vervet monkeys . One infected tsetse fly was allowed to feed on each host individual , and in all cases infections were successful . The characteristics of infection and disease were similar in all hosts , but the rate of progression varied considerably . Parasites were first detected in the blood 4–10 days after infection , showing that migration of parasites from the site of fly bite was very rapid . Anaemia was a key feature of disease , with a reduction in the numbers and average size of red blood cells and associated decline in numbers of platelets and white blood cells . One to six weeks after infection , parasites were observed in the cerebrospinal fluid ( CSF ) , indicating that they had moved from the blood into the brain; this was associated with a white cell infiltration . This study shows that fly-transmitted infection in vervets accurately mimics human disease and provides a robust model to understand better how sleeping sickness develops . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Discussion"
]
| [
"infectious",
"diseases/infectious",
"diseases",
"of",
"the",
"nervous",
"system",
"infectious",
"diseases",
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"pathology/hematology",
"infectious",
"diseases/protozoal",
"infections"
]
| 2008 | Trypanosoma brucei rhodesiense Transmitted by a Single Tsetse Fly Bite in Vervet Monkeys as a Model of Human African Trypanosomiasis |
Pathogenic bacteria use interconnected multi-layered regulatory networks , such as quorum sensing ( QS ) networks to sense and respond to environmental cues and external and internal bacterial cell signals , and thereby adapt to and exploit target hosts . Despite the many advances that have been made in understanding QS regulation , little is known regarding how these inputs are integrated and processed in the context of multi-layered QS regulatory networks . Here we report the examination of the Pseudomonas aeruginosa QS 4-hydroxy-2-alkylquinolines ( HAQs ) MvfR regulatory network and determination of its interaction with the QS acyl-homoserine-lactone ( AHL ) RhlR network . The aim of this work was to elucidate paradigmatically the complex relationships between multi-layered regulatory QS circuitries , their signaling molecules , and the environmental cues to which they respond . Our findings revealed positive and negative homeostatic regulatory loops that fine-tune the MvfR regulon via a multi-layered dependent homeostatic regulation of the cell-cell signaling molecules PQS and HHQ , and interplay between these molecules and iron . We discovered that the MvfR regulon component PqsE is a key mediator in orchestrating this homeostatic regulation , and in establishing a connection to the QS rhlR system in cooperation with RhlR . Our results show that P . aeruginosa modulates the intensity of its virulence response , at least in part , through this multi-layered interplay . Our findings underscore the importance of the homeostatic interplay that balances competition within and between QS systems via cell-cell signaling molecules and environmental cues in the control of virulence gene expression . Elucidation of the fine-tuning of this complex relationship offers novel insights into the regulation of these systems and may inform strategies designed to limit infections caused by P . aeruginosa and related human pathogens .
Microbes translate environmental cues to coordinate and modulate gene expression such that they can adapt to different niches and overcome hostile environments . Adaptation and coordination of gene expression is particularly important for pathogenic microorganisms that need to colonize dynamic host environments since their ability to sense and respond to host environmental cues is critical for their survival . In bacteria , modulation and coordination of gene expression are also influenced by population density via the regulated production of small molecules that serve as intricate signals impacting the expression of virulence factor genes . Many studies have addressed the role of quorum sensing ( QS ) communication networks in virulence where by diffusible intercellular auto-inducers factor and environmental signals bacterial cultures mediate pathogenicity by coordinating the expression of a large array of genes [1] , [2] . Nevertheless , less is known regarding how environmental cues are translated in the context of QS signaling and how environmental cues and QS are integrated to promote the ability of a pathogen to survive and colonize particular niches within their host environments . The processing and integration of environmental inputs in QS becomes even more complex when a pathogen is able to occupy more than one niche . Pseudomonas aeruginosa is a ubiquitous and an extremely versatile Gram-negative bacterium with an astounding ability to survive in many different environments and to infect multiple hosts ranging from amoebas to humans [3] . This pathogen has an extensively studied complex QS communication network that facilitates cross-talk between organisms and impacts many P . aeruginosa group-related behaviors including virulence [4] , [5] , [6] , [7] , [8] , [and 9] . There are at least three known QS systems in P . aeruginosa: two are dependent on the acyl-homoserine-lactone ( AHL ) QS transcription factors LasR and RhlR [10] and a third is dependent on the 4-hydroxy-2-alkylquinolines ( HAQs ) LysR-type transcription factor MvfR [11] , [12] . MvfR activation is mediated by the cell-cell signaling molecules 4-hydroxy-2-heptylquinoline ( HHQ ) and 3 , 4-dihydroxy-2-heptylquinoline ( PQS ) , and leads to the positive regulation of many virulence-related factors , a large number of which are also controlled by the QS signal acyl-homoserine-lactone ( AHL ) -mediated RhlR and LasR circuitry . The MvfR pathway is a critical virulence component essential for the full virulence of P . aeruginosa in multiple hosts [13] , [14] , [15] and is connected to LasR and RhlR by: ( i ) the dependence of mvfR expression at the early growth stages as a result of positive control by LasR [16] , ( ii ) the conversion of HHQ into PQS controlled by PqsH [17] , [18] whose expression is mediated by LasR [19] , [20] , and ( iii ) the negative effects of RhlR on the pqs operon [16] , [21] , which is responsible for the synthesis of all HAQs [11] , [14] , [19] , [22] , [23] including the MvfR ligands HHQ and PQS [12] , [17] , [21] . The QS regulons MvfR , LasR and RhlR respond not only to QS signal molecules but also to environmental signals [24] , including host factors [25] , [26] , [27] , [28] and other environmental cues such as phosphate [29] , magnesium [30] and iron [31] , [32] , [33] , [34] , [35] . Iron acquisition is controlled by a large set of P . aeruginosa genes activated in response to iron starvation [36] , [37] , [38] , including two siderophore complexes , pyoverdine and pyochelin [39] , [40] , and several ferric uptake regulators , among them are the general iron uptake regulator Fur , Fur-regulated pyoverdine siderophore-specific extracytoplasmic sigma factor PvdS , several ECF sigma factors , and the AraC regulator PchR , which regulates pyochelin uptake [40] . In low iron conditions , PvdS binds to iron-starvation ( IS ) boxes to induce the transcription of many genes involved in the iron starvation response [41] . The intricate relationship between QS and iron is exemplified by a series of findings demonstrating that iron starvation induced QS systems [26] , [32] , [34] and that the QS regulators MvfR [11] , LasR/RhlR [42] and VqsR [31] , [43] , [44] were found to be responsible for the induction of many iron response genes . Moreover , MvfR contains an IS box in its promoter [36] , and PQS production is positively-affected by two Fur-regulated small RNAs , Prrf 1 and 2 [45] . Adding to the complexity of how environmental cues such as iron levels affect QS and how iron is integrated into QS to modulate virulence gene expression is the ability of PQS to bind iron [46] , to act as an iron trap molecule [47] , and to form a toxic complex against the host [48] . MvfR activation by HHQ and PQS leads to the upregulation of the anthranilic acid ( AA ) - biosynthetic encoding genes phnAB , and pqsA-E operon [11] , [12] , [14] that have a conserved genomic organization in P . aeruginosa and in HAQs-producing Burkholderia species [49] , to produce more HAQs leading to the upregulation of the MvfR-regulon in a positive feedback loop . Although the fifth gene of the pqs operon pqsE ( PA14_51380 ) , which encodes a predicted GloB , Zn-dependent hydrolase [50] and member of the metallo-beta-lactamase super family ( Pfam PF00753 ) , is not required for HAQ synthesis [12] , [19] , it is co-regulated together with the pqsA-D genes . We have shown that PqsE is essential for complete P . aeruginosa virulence in mice because it controls the expression of a number of MvfR regulon-dependent genes [11] . Although PqsE was previously implicated as the PQS response gene [19] , [20] , it was recently shown to act independently of MvfR and PQS [51] . Thus , the PqsE functions associated with the integration and translation of the QS cell-cell signals has yet to be resolved . Here we examine the interplay between environmental cues and cell-cell signaling molecules and assess how they are integrated in the modulation of MvfR regulon gene expression . To elucidate the QS multi-layered regulation , we also examine the functional dependency of the MvfR regulon components , especially PqsE , and PQS and HHQ , on the Rhl regulon . The findings presented offer new insights into the highly complex P . aeruginosa virulence-associated regulatory loops that may aid in understanding and controlling its pathogenicity .
To elucidate how multi-layered regulatory networks sense and respond to external and internal cell signals to modulate gene expression , we studied the role of MvfR pathway components in integrating and translating signals from PQS and HHQ in the activation of the MvfR regulon genes . To this end , we measured pyocyanin production as an index . This secreted P . aeruginosa phenazine was chosen since its production is dependent on the MvfR pathway components , including the cell-cell signaling molecules , PQS and HHQ , and their corresponding biosynthetic enzymes PqsA-D , their AA precursor , PqsE , and on its Phz biosynthetic operons ( Figure 1A and [11] ) . Here we found that overexpression of PqsE under a constitutive promoter ( pDN19pqsE ) in pqsA− and mvfR− mutant cells not producing HAQs restored pyocyanin production ( Figure 1A ) . In contrast , overexpression of mvfR under a constitutive promoter in a pqsE− background did not restore pyocyanin production ( Figure 1A ) even when HHQ , PQS , or PA14 cell-free supernatants were added ( data not shown ) . These results highlight the crucial role of PqsE in the regulation of MvfR regulon-dependent factors and demonstrate that PqsE possesses activation properties that are independent of HAQ-mediated signals ( Table S1 ) . To assess PqsE mode of action on pyocyanin production , we co-cultured pqsE− cells constitutively expressing the phenazine biosynthetic operon phzA2-G2 with pqsE− cells harboring the phzM and phzS genes essential to pyocyanin synthesis [52] and assessed pyocyanin production . As shown in Figure 1B , approximately 60% of the pyocyanin production was restored , indicating that PqsE participated in pyocyanin production regulation rather than in its synthesis . Second , we tested whether the precursor of all HAQs , AA was required for PqsE function instead . To this end we used a triple mutant strain deficient in phnAB , trpE and kynBU ( AA− mutant ) unable to produce any AA since all three AA synthesis pathways were knocked out [53] . Expression of PqsE in this triple mutant also resulted in high levels of pyocyanin production ( Figure 1A ) corroborating with the above results and demonstrating that PqsE function did not require AA or any of its derivatives to promote production of the MvfR regulon-dependent factor pyocyanin . Third , since PqsE controlled the regulation of one of the key MvfR-regulated factors , pyocyanin , we sought to define the impact of this factor in the regulation of all MvfR-dependent virulence genes . We carried out whole genome expression studies and compared the expression profiles of a pqsE− mutant to those of the PA14 parental strain , an mvfR− mutant and to those of PA14 and an mvfR− over-expressing pqsE strain ( NCBI GEO , accession number #GSE17147 ) . These results showed that PqsE profoundly affected the expression of 90% of the MvfR-regulated genes , including at least thirty-six known and predicted transcription factors ( Tables S1B and S2 ) . Of the PqsE-dependent genes , 241 were found to be negatively regulated and 384 positively regulated by PqsE ( Table S1 ) . At least 75 positively-regulated genes encoded for putative or known virulence factors ( Table S1 ) [11] , [42] . Importantly , included among the positively-regulated virulence transcriptional factors was the QS AHL regulator rhlR [38] and iron response genes , including the iron starvation sigma factor pvdS and genes involved in the synthesis of the siderophore complex pyochelin ( Table S3A ) . To confirm that PqsE overexpression also restores virulence functions apart from restoring their expression independently of the signaling molecules PQS and HHQ , we used two assays . The first is based on the observation that virulent P . aeruginosa strains; including PA14 kill yeast [54] , [55] , [56]; and the second is based on that P . aeruginosa can infect and kill Drosophila melanogaster [57] , [58] , [59] , and that mvfR mutant cells exhibit attenuated virulence in flies [57] . As illustrated in Figure 1C–D , a zone of yeast growth inhibition was observed around PA14 , but not around the mvfR− , or pqsE− mutants following plating of C . neoformans KN99α 5 mm from the bacterial colony on a YPD plate ( Figure 1D ) . The killing zone was restored following PqsE overexpression in mvfR− backgrounds ( Figure 1C–D ) . In agreement flies infected with pqsA− or pqsE− mutants cells exhibited significant delayed in mortality compared to that caused by the WT or the pqsA− cells expressing pqsE ( Figure 1E ) demonstrating again that PqsE is crucial for P . aeruginosa pathogenicity and independent of PQS and HHQ . Comparison of the pqsE transcriptome ( Table S1 ) to lasR/rhlR [42] revealed that almost half ( 46% ) of the genes regulated by LasR/RhlR were also regulated by PqsE ( Figure S3A ) indicating a relationship between AHL- and MvfR-mediated QS regulons . This relationship is also extended to the negative effects that both components have on the transcription of the pqs operon ( [16] and Table S1 and Figure 2A ) . A green fluorescent protein ( GFP ) reporter gene [32] fused to the pqs operon promoter ( Figures 2B ) , quantitative PCR analysis ( Figure S2D ) and quantification of HHQ and PQS levels ( Figure 2C ) further validated the above finding . Moreover , in agreement , Figure 2D shows that HAQ synthesis down-regulation paralleled the accumulation of AA ( HAQ precursor ) followed by an increase in antABC gene expression that encodes enzymes for AA degradation ( Table S1 ) . To determine whether there was indeed a functional relationship between the respective communication-systems components RhlR and PqsE in the regulation of the MvfR regulon signal production and whether they together affected signal integration , we proceeded to assess whether there was a RhlR-PqsE codependency in the negative regulation of HAQ biosynthesis . Figures 3A and S4B show that overexpression of PqsE in a rhlR− mutant did not result in a downregulation of the promoter-derived expression of the pqs operon in contrast to the overexpression of PqsE in the wild-type ( WT ) strain PA14 where expression of the pqs operon was downregulated ( Figure 2 and Figure S2D ) . These results indicate that PqsE negative control of the activity of the MvfR regulon depends on RhlR . Second , we examined whether there was an RhlR-PqsE codependency in signal integration by MvfR-regulon virulence genes downstream of PqsE . To this end , we assessed whether PqsE overproduction in rhlR− cells could restore pyocyanin production since it was completely abolished in both pqsE− [11] , [19] and rhlR− [38] mutants . Figure 3B shows that PqsE did not restore pyocyanin production in rhlR− while RhlR expression partially ( ∼30% ) restored pyocyanin production in pqsE− mutant cells . This finding suggests that PqsE also depends on RhlR in the positive regulation of pyocyanin production and that RhlR acts downstream of PqsE . Interestingly , Figure S5 shows that pyoverdine levels are higher in rhlR− than in PA14 but not in pqsE− mutant cells . Moreover , PqsE or RhlR overproduction in rhlR− or pqsE− mutant cells respectively did not fully downregulated pyoverdine production , while PqsE or RhlR overproduction in the corresponding mutant cells did ( Figure S5 ) . This finding suggests RhlR-PqsE codependency in the homeostatic regulation of pyoverdine . Based on the above findings , it is likely that the PqsE-RhlR activities were not limited to controlling downstream genes associated only with pyocyanin or pyoverdine production if the high number of genes co-regulated by PqsE and the Las/Rhl system are considered ( Figure S3A ) . The pyocyanin levels produced by the non-HAQs producing mutants pqsA− , mvfR− and AA− [12] , [19] , [22] , [53] overexpressing pqsE were higher than the levels produced by the HAQs-producing PA14 parental strain carrying the same plasmid ( Figure 1A ) . This difference raised the question regarding whether the presence and/or levels of HAQs had dose-dependent negative effects on pyocyanin levels . To this end we assessed the effect of exogenously-added HAQs on pyocyanin levels by using 20 mg/L of PQS or HHQ , a concentration corresponding to the approximate maximal physiologic levels reached by PA14 or pqsH− strains respectively at stationary phase ( [17] and Figure S4A ) . Figure 4A shows that the pyocyanin levels in either pqsA::pqsH− or mvfR− mutants overexpressing pqsE were significantly lower in the presence of either HHQ or PQS . Figure 4B shows that PQS concentrations ( up to 1 mg/L ) induced pyocyanin production in both pqsH− and pqsA−::pqsH− cells but concentrations >1 mg/L decreased pyocyanin production in a dose-dependent manner in all strains tested ( Figure 4B ) without significantly affecting cell growth ( data not shown ) . This concentration-dependent decrease in pyocyanin levels was independent of PqsE function and phz operon regulation since it was also observed in pqsE− cells constitutively expressing phz genes ( Figure 4C ) . The PQS-mediated down-regulation was not specific to PA14 cells as it was also observed in the PA01 P . aeruginosa strain ( Figure 4C ) . To determine whether high physiological levels of PQS and/or HHQ negatively-impact pqs operon gene expression , we conducted experiments using pqsA−::pqsH− cells harboring the pqsA-GFP ( ASV ) reporter gene . Figure 4D shows that 20 mg/L HHQ negatively-impacted pqsA gene expression compared to 10 mg/L . PqsA gene expression was not affected by any of the PQS concentrations tested . Interestingly , a negative effect on pqsA gene expression , similar to that observed following treatment with 20 mg/L HHQ , was also observed when the two HAQs were added together in sub-inhibitory concentrations ( 1 mg/L PQS +10 mg/L of HHQ ) . This result is indicating that together HHQ and PQS have synergistic inhibitory effect and implying also that high activation of the pqs operon led to its down-regulation . To further elucidate the role of PQS on PqsE-dependent gene regulation , we compared the transcriptional profiles of mvfR− mutant cells overexpressing PqsE in the absence or presence of 20 mg/L PQS ( Table S1 ) . High PQS concentrations negatively affected the expression of 191 of 625 ( 31% ) PqsE-regulated genes ( Figure 4E and Table S1 ) . This effect was more apparent among the known and putative virulence factors where the expression of 64% of the PqsE-regulated genes , ( including chitinase , halovibrin , cellulase , pyocins , lectin , and elastase genes ) was significantly reduced by more than 2-fold upon PQS addition ( Table S1 ) . The addition of PQS further increased the expression of only 7 genes: fpvA , the major pyoverdine receptor; gatC , a Glu-tRNA amidotransferase subunit C; sucA , a 2-oxoglutarate dehydrogenase; bkdA1 , a 2-oxoisovalerate dehydrogenase and of three hypothetical proteins; PA4642 , PA1343 and PA2405 ( Table S1 ) . Interestingly , transcription of phz operon genes was not modified by the addition of PQS although pyocyanin production was affected ( Figure 4A ) , suggesting that PQS may be acting post-transcriptionally in this case . As shown in Table S3A , PqsE positively-affected the expression of 43 iron starvation-related genes [36] , [37] including the iron starvation sigma factor PvdS [41] , [60] , the pyochelin regulator PchR [61] , vqsR [31] , [62] and PA2384 [63] . Interestingly , PqsE negatively regulated only 6 iron related genes , bfrB and the siderophore pyoverdine associated genes pvdA pvdF , pvdJ , pvdN and pvdQ ( Table S3A ) reflected also in the pyoverdine levels ( Figure S5 ) . It is noteworthy that PqsE acted differentially on the siderophores , serving as a positive regulator of pyochelin and a negative regulator of pyoverdine ( Figure S5 ) . In addition , Table S3A reveal that HAQs are also involved in the control of iron-related genes by PqsE since constitutive expression of pqsE triggered this effect in the mvfR− background cells lacking HAQs but not in PA14 cells . To examine how iron starvation is translated in the context of MvfR signaling , we first examined whether there is a relationship between iron starvation and the regulation of PQS and MvfR regulon genes . We compared pqsA transcription using a pqsA-GFP ( ASV ) reporter in PA14 cells grown in the absence ( D-TSB medium ) or presence of high iron levels . Figure 5A demonstrates that iron significantly reduced pqsA transcription . Subsequently , we examined the effect of iron directly on the induction of pqs operon transcription in presence only of PQS and not of other HAQs in pqsA−::pqsH− mutant cells . Using 1 mg/L PQS , an amount sufficient to fully induce pqs operon transcription and increasing concentrations of FeCl3 Figure 5B shows an iron concentration-dependent effect on pqsA gene expression . We next examined if iron could also counterbalance the downstream effects of PQS on PqsE-dependent genes by assessing the effect of HAQs and iron on pyocyanin production . Figure 5C shows that the addition of iron abolished the reduction in pyocyanin production conferred by PQS ( 20 mg/L ) and restored pyocyanin production to that observed in the presence of 1 mg/L PQS . A similar effect was observed in PA14 cells and pqsA−::pqsH− cells overexpressing PqsE ( Figure S6A ) where the addition of 20 mg/L PQS decreased pyocyanin levels which were restored in the presence of iron . Since iron alone did not affect pyocyanin production in the experimental conditions used , it suggested that pyocyanin production was affected due to direct effect of iron on PQS . No significant difference in growth was observed between PA14 cells grown in absence or presence of various concentrations of iron ( up to 250 µM , Figure S6B ) . Collectively , these findings indicate that iron counterbalanced PQS-dependent regulation by ‘fine-tuning’ its activity , possibly by reducing PQS activity when it is in a complex with it .
In this work , we delineated paradigmatically the complex relationships between bacterial multi-layered regulatory QS circuitries , their signaling molecules , and the environmental cues to which they respond . The intracellular communication system of P . aeruginosa possesses complex signal transduction systems allowing this versatile pathogen to regulate and coordinate virulence functions in the context of multiple hosts , environments , and competition from other microorganisms [7] , [64] , [65] , [66] . Here we showed that one of these complex signal transduction systems , MvfR , responds to both positive and negative feedback loops that are interconnected with the RhlR QS complex system and that these interactions fine tune the production and concentration of secreted output signals that in turn serve as inputs to preserve a homeostatic regulation . Moreover , our experiments demonstrated that via the finely tuned cooperation and homeostatic interplay between the MvfR circuitry components PqsE , and PQS and HHQ with RhlR and iron , this pathogen governs and balances the intensity of its virulence response . Although HHQ and PQS principally serve as MvfR ligands [17] , [18] , our results show that once maximal in vitro physiological levels are reached , they negatively impact their own production and the downstream PqsE regulated genes . PqsE , HHQ and PQS are essential molecules in the negative feedback auto-regulatory loops that contribute to this homeostatic regulation . Although the HHQ concentrations shown here are not attained in vitro because HHQ is fully converted into PQS , this effect is most likely relevant in vivo where we have shown that HHQ levels are higher than those of PQS [17] . In addition , in lasR− mutants that accumulated during chronic infections HHQ levels are also higher than PQS since PqsH responsible for the conversion of HHQ to PQS is under the control of LasR [67] . Nevertheless , we show that HHQ and PQS have together synergistic effect as a negative auto-regulators that down-regulated pqs operon transcription , reducing their own production and that of the other HAQs . Thus , jointly with PqsE , PQS and HHQ most probably contributed to the down-regulation of the pqs and phn operons observed during the late growth phase of P . aeruginosa ( Figure S1 ) . In addition to being activator and auto-down-regulator PQS acted also as a homeostatic agent at high physiological concentrations by down-regulating most PqsE-dependent , downstream genes . Consistently , maximum pyocyanin production occurred only at low PQS concentrations that were sufficient to maximally activate the pqs operon . The homeostatic effect of PQS downstream of the PqsE genes was clearly independent of MvfR , PqsE and of other HAQs given that its effects were still apparent in pqsA− and mvfR− backgrounds . Interestingly , this effect appeared also to be post-transcriptional since PQS did not significantly impact phz operon transcription but affected pyocyanin production even when the phz operon was constitutively expressed . The mechanism behind this effect remains to be discovered . One intriguing possibility may be that PQS exerts its effect via RsmA and/or on small RNAs like rsmZ or prrF . Previous studies have suggested that while PqsE is the PQS response protein [19] , [20] , it does not influence PQS production [11] , [12] . Here we show that PqsE is a crucial player in orchestrating the homeostatic regulation of the signaling molecules HHQ and PQS as well as establishing a connection to the QS RhlR system , underscoring it as a key mediator of MvfR regulon activation and cooperation with the AHL QS system . Our findings also provide initial answers as to why PqsE , although not involved in the synthesis of HAQs in vivo or in vitro [11] , [19] , [20] ) , is tightly regulated together with the other pqs operon genes . Although our findings are primarily based on trans-regulatory studies , the overexpression of PqsE demonstrated for the first time that PqsE can impact HAQs concentrations by down-regulating their production . In corroboration , are both the AA accumulation and the transcriptional induction of the antABC genes responsible for AA degradation [68] , [69] and shown to be regulated by prrF1 and prrF2 [45] . Since pqsE is co-transcribed by MvfR together with pqsA-D genes , the reduced production of HAQs mediated by PqsE indicates that pqsE gene transcription itself is also downregulated in a negative feedback mechanism that finely balances the regulatory loop . Although PQS and HHQ signal molecules are critical to MvfR-dependent gene expression , their addition has failed to rescue pqsE- mutant cells to activate expression of many MvfR-regulated genes or to produce of pyocyanin [11] , [17] , [19] , [20] . Here we found that overexpression of PqsE induced pyocyanin production and transcription of an additional approximately 600 MvfR-regulated genes independently of MvfR , HAQs and AA , demonstrating the crucial role of PqsE in activating MvfR regulon genes independently of the HAQs . Ultimately , expression of PqsE in an mvfR− or pqsA− strain restored P . aeruginosa virulence as determined by growth inhibition of yeast and flies feeding assay , indicating that PqsE did not need HAQs to confer virulence in these systems . Corroboratory results were reported by Farrow et al . [51] who showed in a qualitative manner that expression of PqsE in an mvfR− mutant restored pyocyanin production . These results together indicate that , at least with regard to the genes listed in Table S1 , PQS and HHQ only act as inducers of MvfR to express PqsE that once expressed induces the P . aeruginosa virulence response without HAQs or MvfR . Thus , PqsE cannot be designated as the “quinolone signal response protein” . Nevertheless , it is not yet known how PqsE , a protein that belongs to the metallo-beta-lactamase super family without any known DNA binding motifs , regulates the transcription of so many genes . Its predicted hydrolase activity suggests that it may cleave or participate in the synthesis of small molecules . Due to the location of the pqsE gene in the pqs operon , the immediate candidates likely targeted by PqsE are HAQs . However following extensive LC/MS analyses , we were unable to detect any molecule that accumulated or diminished in concentration in pqsE− cultures compared to WT cultures ( data not shown ) . In addition we were unable to complement pyocyanin production in a pqsE− culture by exogenously adding HAQs , AHLs or whole PA14 supernatants ( [11] and data not shown ) . Nonetheless , collectively , our results indicate that PqsE is involved in a negative feedback loop that affects the regulation and integration of HAQs-mediated cell-cell signaling molecules and that is functionally dependent on RhlR . The exact nature of the co-dependency between PqsE and RhlR remains unclear . The downregulation of rhlR expression by ∼2 fold in a pqsE mutant is not sufficient to explain the striking transcriptional and phenotypic effects mediated by PqsE . Since PqsE is not predicted to be a transcriptional factor [50] it is highly likely that it may exert its effect on RhlR post-transcriptionally , and this effect may be perhaps extended to other transcriptional factors . The MvfR affected gene list has a substantial overlap [11] with the previously published list of Rhl/Las-controlled genes [42] , and the expression of almost all MvfR-regulated genes controlled by PqsE . Both PqsE activities ( i . e . , fine-tuning HAQs production by down-regulating the pqs operon , induction of pyocyanin production and downregulation of pyoverdine production ) were dependent on RhlR apparently acting downstream but in a tight collaboration with PqsE . Recently , Farrow and colleagues showed that the addition of AHL C4-HSL ( a RhlR inducer ) to PAO1 pqsE− isogenic mutants also restored pyocyanin production [51] . These findings , although we did not reproduce them in PA14 cells , are in agreement with our findings that PqsE and RhlR functions are linked . However , the exact relationship between PqsE and RhlR , that is when or how they cooperate , remains elusive since RhlR in some cases functions in the absence of PqsE; for example , the RhlR-dependent C4-HSL levels in a pqsE− mutant strain were identical to the parental strain ( data not shown ) as also was previously shown for the mvfR− mutant [11] . The relationship between iron , QS regulation , and P . aeruginosa virulence is multifaceted [31] , [32] , [34] , [36] , [45] , [63] and extremely complex . Data presented in this report demonstrate that the MvfR regulon represents a striking paradigm of the interplay between environmental signals and bacterial secreted cell-cell signal molecules that participate in positive and negative homeostatic regulatory loops . QS MvfR components control the transcription of many iron related genes , while iron related regulators control the expression of QS genes ( see Table S3B ) in addition to iron related genes . The relationship between iron and QS regulation is further strengthened through the iron-related regulators VqsR [43] and the PA2384 product [63] that were found to control the expression of phnAB and pqsA-E operons . Furthermore , the iron starvation sigma factor PvdS was shown to positively control the expression of mvfR via its IS box [36] , iron was shown to control the pqs operon during biofilm formation [32] , and the two small Fur-regulated RNAs Prrf 1 and 2 positively-regulated PQS production [45] . Our results showing that iron levels affected HAQs activities both as inducers of MvfR and as fine-balancers provide corroboration for the view that the MvfR regulon is closely linked with iron regulation . The complexity of the interplay between the MvfR regulon and iron control is further increased by: a . the ability of PQS but not HHQ to trap iron [47] , which likely reduces available iron within the cell and promotes iron starvation , thereby affecting PqsE-mediated control of bacterial iron response genes , including the siderophores pyochelin and pyoverdine; and b . iron , especially in high concentrations , induces oxidative stress that was shown to affect and being affected by PQS [70] . Thus , it is possible that some of the phenotypic effects of PQS and iron shown here could be attributed to oxidative stress . Thus , it would be of importance to further investigate the contribution of iron , as a nutrient , a signal molecule , and an oxidative stress inducer in QS and P . aeruginosa virulence . The existence of a tight interconnection between iron concentrations , QS , and virulence in P . aeruginosa is likely due to iron conditions encountered in vivo [71] , [72] serving as a signal indicating a hostile environment requiring expression of virulence or fitness-related genes . When host tissues become damaged as a consequence of virulence factor production , the resulting increase in iron concentrations should down-regulate virulence factor concentrations , thereby reducing bacterial virulence that may favor host survival and potentially chronic infection . A complete understanding of the regulation of the multiple P . aeruginosa virulence networks , in particular the mechanisms of the homeostatic and down-regulation processes ( Figure 6 ) , will be essential for the development of drugs targeting QS inhibition [73] , [74] . The findings presented in this study may aid in the design of anti-infective therapies tailored to interfere with virulence pathways and provide a paradigm for understanding the complex QS networks of other bacterial pathogens besides that of P . aeruginosa .
Table S4 lists bacterial strains and plasmids used in this study . P . aeruginosa were routinely grown in Luria Bertani ( LB ) broth at 37°C for 18 h , and diluted to OD600 nm 0 . 05 and grown to the desired OD600 nm . For low iron media the bacteria were grown in D-TSB medium [36] that was treated with Chelex 100 beads ( Bio-Rad , Hercules , CA ) and for high iron FeCl3 or FeSO4 were added at concentrations of 200 µM . The E . coli JM109 strain was used for sub-cloning and plasmid propagation . The E . coli S17-1 strain was used for conjugation between E . coli and P . aeruginosa by the pEX18Ap-derivative allelic replacement method [75] . Antibiotics used included ampicillin ( Amp ) ( 100 µg/ml ) , carbenicillin ( Crb ) ( 300 µg/ml ) , gentamycin ( Gnt ) ( 15/60 µg/ml ) , kanamycin ( ( Kan ) , 50/200 ) , tetracycline ( Tet ) ( 15/200 µg/ml ) and chloramphenicol ( Cam ) ( 15/50 µg/ml ) for E . coli and P . aeruginosa respectively . The plasmid overexpressing PqsE was generated by PCR amplification of the pqsE gene from PA14 genomic DNA using primer pairs GX119 and GX120 ( Table S4 ) . The PCR product was digested with HindIII/XbaI and sub-cloned into the pDN19 plasmid vector under plac promoter to generate pDN19pqsE that constitutively expresses pqsE . Construct integrity was confirmed by DNA sequencing . Plasmids were introduced into E . coli or P . aeruginosa PA14 by electroporation . Non polar deletions were generated by pEX18AP allelic replacement using sucrose selection . Fragments with the size of about 1 kb flanking the desired genes were cloned into the pEX18Ap plasmid vector and introduced into E . coli by electroporation followed by conjugation to P . aeruginosa . Alternatively , the λ-Red recombinase method was used to generate chromosomal deletions or insertions [53] . Two kinds of reporter genes were used: 1 ) translational and transcriptional fusions to lacZ where the β-galactosidase activity assay was performed in triplicate as described [76] and; 2 ) a pqsA-GFP ( ASV ) fusion consisting of a pqsA promoter upstream to a short-lived GFP that allows for the detection of pqs operon up or down regulation carried on the plasmid pAC37 [32] . Overnight cultures were diluted to an OD600 nm of 0 . 05 in black , clear bottom sterile 96-well assay plates ( Corning Inc . , Corning , NY ) . The plates were incubated for 50 h at 37°C in an Infinite F200 plate reader ( Tecan Group Ltd , Männedorf , Switzerland ) . Every 30 min the plates were shaken for 2 min and read at 600 nm and fluorescence detected by excitation at 485 nm and emission at 535 nm . The results are expressed as an average of 3–6 observations that were normalized to a strain that did not carry the plasmid pAC37 . Bacteria were respectively grown overnight at 37°C , diluted to an OD600 nm of 0 . 05 in 25 ml LB with the corresponding antibiotics at 37°C until the OD600 nm reached 3 . 0 . The total RNA was isolated with the RNeasy Mini kit ( QIAGEN Inc . , Valencia , CA ) and cDNA synthesis and labeling performed according to the manufacturer's instructions ( Affymetrix , Santa Clara , CA ) . The P . aeruginosa PAO1 GeneChip® Genome array ( Affymetrix ) was used for hybridization , staining , washing and scanning according to the manufacturer's instructions . Experiments were independently performed in triplicate . Affymetrix DAT files were processed using the Affymetrix Gene Chip Operating System ( GCOS ) to create . cel files . The raw intensity . cel files were normalized by robust multi-chip analysis ( RMA ) ( Bioconductor release 1 . 7 ) with PM-only models . Array quality control metrics generated by the Affymetrix Microarray Suite 5 . 0 were used to assess hybridization quality . Normalized expression values were analyzed with SAM ( Significance Analysis of Microarray ) using the permuted unpaired two-class test . Genes whose transcript levels exhibited either a 2-fold or up or down regulation and had a q value <6% were further analyzed . The results of the GeneChip® arrays were imported to GeneSpring 7 . 3 ( Agilent Technologies , Inc . , Palo Alto , CA ) and the expression signals of the GeneChip® arrays were normalized to the constant value of 1 . 0 and the ratio cut-off was set to 2-fold . Annotations were performed using the database http://pseudomonas . com/ . The transcriptome results were ( in part ) validated by assessing β-galactosidase expression and RT-PCR of selected genes ( Figure S2 ) . The data are deposited in NCBI GEO with accession number #GSE17147 . Cells from each triplicate experiment were harvested at an OD600 nm of 2 , 3 and 4 . Total RNA was subsequently isolated using the RiboPure-Bacteria RNA Isolation kit ( Ambion , Austin , TX ) according to the manufacturer's instructions . cDNAs were synthesized with random reverse primers using the Reverse Transcription RETROscript kit ( Ambion ) according to the manufacturer's instructions . Specific primers ( Table S4 ) for the amplification of products of approximately 200 base pairs were designed using the Primer3 algorithm ( http://frodo . wi . mit . edu/primer3/ ) and analyzed by In Silico simulation of PCR amplifications ( http://insilico . ehu . es/ ) and by the Primer Analysis Software NetPrimer ( Premier Biosoft International , http://www . premierbiosoft . com/netprimer/index . html ) for the detection of expressed pqsA , pqsE and rpoD that served as the normalizer genes [77] . Quantitative RT-PCR was carried out using the Brilliant II SYBR Green QPCR Master Mix ( Stratagene ) with a RT Fluorescence Detection System MX3005P ( Stratagene , La Jolla , CA ) in a 25 µl final volume . The efficiency of each pair of primers was determined by a standard curve of 8 dilutions of 1∶4 of PA14 genomic DNA . The relative expression ratios were calculated and analyzed using MXPro analysis software , version 4 . 01 ( Stratagene ) using a mathematical model that included an efficiency correction . The fold induction of mRNA was determined from the threshold values that were first normalized for rpoD gene expression that served as a normalizer and then for the threshold value of the WT strain harboring the pDN19 plasmid at an OD600 nm of 2 that served as the calibrator . The data are expressed as the average of triplicate samples . The quantification of HAQs concentration in bacterial culture supernatants and in vivo from rectus adominus muscle of burned and infected mice was performed by LC/MS as described previously [17] , [78] . The HAQs were separated on a C18 reverse-phase column connected to a mass spectrometer using a water/acetonitrile gradient [78] . Positive electrospray in the MRM mode with 2×10−3 mTorr argon and 30 V as the collision gas were employed to quantify HAQs using the ion transitions HHQ 244>159 , HHQ-D4 248>163 , HQNO 260>159 , PQS 260>175 , and PQS-D4 264>179 . The pseudomolecular ions of each compound were monitored in full scan mode using the unsaturated PA14 HAQs response factors . Samples of 5 ml were spun down and the supernatants mixed with equal volumes of chloroform . The lower blue organic phase was collected and mixed with 5 ml of HCl ( 0 . 2 N ) . The upper reddish phase was collected and its OD52 onm was measured . The concentration of pyocyanin was determined by the formula: mg/L = OD52 onm×17 . 072 normalized to cell counts and the statistical significance was assessed using the Student's 2 tailed t-test assuming equal variance [79] . In order to assess the production of pyocyanin by expression of the phz genes we used a co-culture of cells harboring the pUCP-A2G2 and pUCP-MS plasmids [80] . All experiments were performed in triplicate . D-TSB medium was used to grow 200 µl of bacterial cells in 96 wells plate . Production of pyoverdine was assessed using a plate reader ( Infinite F200 , Tecan Group Ltd , Männedorf , Switzerland ) . Pyoverdine levels were determined every 30 minutes using excitation at 400 nm and emission at 460 nm and the values obtained were normalized to cell growth ( OD600 nm ) . Pyoverdine concentrations were calculated using a calibration curve of fluorescence of a range of concentrations of pyoverdine ( Sigma Aldrich , US ) . Yeast ( Cryptococcus neoformans KN99 α , Candida albicans ATCC #90028 DAY185 strain or Saccharomyces cerevisiae YJM310 strain ) were plated for 2 days on YPD agar ( Difco ) plates at 30°C . A colony was picked and grown for 18 h in liquid YPD media ( Difco ) at 30°C with shaking ( 200 rpm ) . The yeast was diluted 1∶100 in 4 ml soft YPD agar ( 0 . 6% agar ) and poured onto an YPD plate that was dried for 30 min in a laminar flow hood . A 1 µl drop of an overnight culture of the desired bacterial strain was put on top of the yeast lawn and the plate incubated for 2–3 days at 30°C . A dead yeast zone was formed around a by PA14 bacterial colony bun not around mutants such e . g . , mvfR − , pqsA − and pqsE − . The viability of yeast in these zones was tested by plating yeast from distance of 5 mm from the bacterial colonies on YPD plates . Fly infection feeding assay was performed as previously described in [58] , [59] . Briefly , 45 female Oregon-R flies per group , 5–7 days old , were fed with a mixture of 4 ml of LB bacterial culture at OD600 nm 3 . 0 with 1 ml of 20% sucrose . Thus , feeding mix contained a final concentration of 80% LB containing ∼2×109 bacterial cells per ml and 4% sucrose . An autoclaved cotton ball was placed at the bottom of each fly vial and was impregnated with 5 ml of the feeding mix . The 45 flies per treatment group were sub-divided in three fly vials ( 15 flies in each ) , sealed with a clean cotton ball , and incubated at 25°C . Fly survival was recorded twice a day until all flies succumbed to the infection . Statistical analysis of the survival curves was preformed using the log-rank test ( Mantel-Haenszel ) of the Kaplan-Meier estimate of survival using the software MedCalc ( http://www . medcalc . be/ ) . Two independent experiments gave similar results . | Bacterial cells can communicate with one another about their surrounding environment . This information can be in the form of small self-secreted molecules acting as signals to activate or inhibit the expression of genes . Pseudomonas aeruginosa is an environmental bacterium that infects diverse organisms from plants to humans . Our results show that this pathogen uses two highly sensitive networks , namely MvfR and LasR/RhlR pathways , to modulate its virulence functions by titrating the concentration of the small molecules HHQ and PQS in a manner that depends upon the presence or absence of iron . Via negative and positive feedback loops , this bacterium processes the signaled information to regulate its virulence functions and homeostatically balance the production of the small molecules required for the activation of the MvfR virulence network . Our study sheds light on paradigmatic complex networks that maintain a homeostatic bacterial virulence response . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"infectious",
"diseases/nosocomial",
"and",
"healthcare-associated",
"infections",
"infectious",
"diseases",
"microbiology/microbial",
"physiology",
"and",
"metabolism",
"biotechnology/applied",
"microbiology",
"microbiology",
"infectious",
"diseases/bacterial",
"infections",
"microbiology/applied",
"microbiology",
"microbiology/microbial",
"growth",
"and",
"development",
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis",
"microbiology/medical",
"microbiology"
]
| 2010 | Homeostatic Interplay between Bacterial Cell-Cell Signaling and Iron in Virulence |
All herpesviruses have mechanisms for passing through cell junctions , which exclude neutralizing antibodies and offer a clear path to neighboring , uninfected cells . In the case of herpes simplex virus type 1 ( HSV-1 ) , direct cell-to-cell transmission takes place between epithelial cells and sensory neurons , where latency is established . The spreading mechanism is poorly understood , but mutations in four different HSV-1 genes can dysregulate it , causing neighboring cells to fuse to produce syncytia . Because the host proteins involved are largely unknown ( other than the virus entry receptor ) , we were intrigued by an earlier discovery that cells infected with wild-type HSV-1 will form syncytia when treated with salubrinal . A biotinylated derivative of this drug was used to pull down cellular complexes , which were analyzed by mass spectrometry . One candidate was a protein tyrosine phosphatase ( PTP1B ) , and although it ultimately proved not to be the target of salubrinal , it was found to be critical for the mechanism of cell-to-cell spread . In particular , a highly specific inhibitor of PTP1B ( CAS 765317-72-4 ) blocked salubrinal-induced fusion , and by itself resulted in a dramatic reduction in the ability of HSV-1 to spread in the presence of neutralizing antibodies . The importance of this phosphatase was confirmed in the absence of drugs by using PTP1B-/- cells . Importantly , replication assays showed that virus titers were unaffected when PTP1B was inhibited or absent . Only cell-to-cell spread was altered . We also examined the effects of salubrinal and the PTP1B inhibitor on the four Syn mutants of HSV-1 , and strikingly different responses were found . That is , both drugs individually enhanced fusion for some mutants and reduced fusion for others . PTP1B is the first host factor identified to be specifically required for cell-to-cell spread , and it may be a therapeutic target for preventing HSV-1 reactivation disease .
There are two ways that viruses can spread to uninfected cells . Cell-free spread occurs when virions are released from an infected cell into their surrounding environment prior to entering a new cell . This , of course , is how viruses spread to new hosts and often between cells within a host . However , some viruses , including all the herpesviruses , also have a “cell-to-cell” spreading mechanism by which virions pass directly through cell junctions , enabling protection from neutralizing antibodies [1 , 2] . For example , herpes simplex virus type 1 ( HSV-1 ) utilizes cell-to-cell spread to move directly from mucosal epithelial cells , the initial site of infection , into nearby sensory neurons , where the virus establishes a latent infection . When the virus reactivates , newly formed viral particles travel back down the axon , and cell-to-cell spread is used again to allow passage of the virions into the mucosal epithelium [3 , 4] . Importantly , replication-competent mutants of HSV-1 that are defective for cell-to-cell spread fail to infect neurons when tested in animal models , and therefore cannot establish latency [5 , 6] . Despite its importance , the mechanism of cell-to-cell spread remains poorly understood for all herpesviruses . Cell-to-cell spread can be assessed in vitro by measuring the sizes of plaques produced in the presence of neutralizing antibodies . These antibodies will inactivate virions released into the medium , preventing cell-free spread . In the presence of neutralizing antibodies , wild type HSV-1 forms large plaques in cell cultures due to its capacity for cell-to-cell spread [7] . In contrast , replication-competent mutants defective for this spreading mechanism exhibit greatly reduced plaque sizes under these conditions [5 , 7] . Hence , this assay has been used to identify viral factors involved in cell-to-cell spread . The complexity of cell-to-cell spread for HSV-1 is reflected in the many viral proteins that seem to be required . At the core , four glycoproteins—gB , gH/gL , and gD—form the fusion complex ( Fig 1A ) , which is sufficient for virus entry but not for the cell-to-cell spread mechanism [8] . During entry , gD binds to cellular receptors and transmits a signal through the gH/gL heterodimer to the viral fusion protein , gB [9 , 10] . This induces a conformational change in gB , triggering the fusion of the viral envelope with the cell membrane [11] . When genes encoding gD , gH/gL , and gB are co-transfected into cells that express receptors , gB activation also occurs , causing massive fusion of the cells with one another [12] . Since cell fusion rarely occurs in HSV-1 infections , it is clear that additional viral proteins regulate the fusion activity of the core fusion complex . Two viral multi-pass membrane proteins known to interact with gB are gK and UL20 ( Fig 1A ) . These proteins also interact with each other and function in cell-to-cell spread in vitro [13 , 14] . gK also regulates the core fusion complex as it blocks cell fusion when co-expressed with gD , gH/gL , and gB [15] , and in mouse models , gK-deletion mutants fail to spread into neurons after ocular infection [16] . gE also plays a critical role in cell-to-cell spread , as shown by gE-deletion mutants , which exhibit spreading defects in vitro in the presence of neutralizing antibodies and fail to spread into neurons in mouse models [5 , 17] . Furthermore , the tegument proteins UL16 and UL21 and peripheral membrane-binding protein UL11 form a complex on the cytoplasmic tail of gE ( Fig 1A ) . In cell cultures , this complex seems to be important for cell-to-cell spread [7 , 18] . Other HSV-1 proteins that have been implicated in cell-to-cell spread include UL51 , gI , and UL34 [19–21] . Strikingly , many of the accessory proteins needed for cell-to-cell spread are also necessary for the syncytial ( Syn ) phenotype exhibited by certain HSV-1 mutants . These mutants inappropriately cause cell fusion , resulting in large multinucleated cells . Most syncytial mutations ( Fig 1A ) cause changes in the cytoplasmic tail of gB or the amino-terminal segment of gK , but alterations in UL20 and UL24 can also produce the Syn phenotype [22] . The Syn phenotype requires the core-fusion complex and a variety of accessory proteins . This has been explored in greatest depth for gBsyn mutants , where removal of gE , UL16 , UL11 , or UL21 results in loss of the Syn phenotype even though the viruses still replicate [18] . Because their requirements clearly overlap , studies of the Syn mutants of HSV-1 have the potential to reveal mechanistic insights for cell-to-cell spread . Although much is known about the viral machinery involved in cell-to-cell spread , virtually nothing is known about required host factors , other than the need for gD to interact with a receptor [23] . A potential way forward emerged 10 years ago with a report showing that wild-type HSV-1 will cause cells to fuse into massive syncytia when they are treated with the drug salubrinal [24] . We reasoned that the cellular target of salubrinal may be involved in regulating both syncytia formation and cell-to-cell spread . Salubrinal is well known to prolong the survival of cells experiencing ER stress , which arises when proteins are made faster than they can be folded [25] . The ER stress response , also called the unfolded protein response ( UPR ) , causes the phosphorylation of translation initiation factor eIF2α , which slows translation and allows time for recovery . Later , GADD34 binds to protein phosphatase 1 ( PP1 ) and directs it to eIF2α to remove the inhibitory phosphate [26] . Salubrinal blocks this step , prolonging the stress response , but its cellular target is unknown [27] . Infection with HSV-1 also triggers the ER stress response , but to keep translation going , the virus encodes ICP34 . 5 , a GADD34 homologue , to recruit PP1 and dephosphorylate eIF2α [28] . Salubrinal blocks this step , too , which extends the translation block , thereby reducing the production of infectious virions [25] . For unknown reasons , salubrinal treatment also causes widespread fusion of HSV-1 infected cells , even in the absence of a Syn mutation or ICP 34 . 5 [24] . We began this study of salubrinal-induced fusion of HSV-1-infected cells with hopes of discovering a host factor involved in the biologically and clinically important process of cell-to-cell spread . In this study , we show that PTP1B , a host tyrosine phosphatase , though not a target of salubrinal , is critical for this type of spread .
To confirm that salubrinal stimulates fusion , Vero cells were infected with the wild-type KOS strain of HSV-1 and incubated with increasing concentrations of the drug . The cells were fixed at 18 hours post infection , and the nuclei were stained with DAPI while a tight junction protein , ZO-1 , was stained with a fluorescent antibody to help identify the plasma membrane . Only HSV-infected cells treated with salubrinal formed syncytia ( Fig 1B ) . Manual counting of all nuclei within syncytia ( defined as 3 or more nuclei per cell ) versus all cells with single nuclei revealed a dramatic increase in fusion as the drug concentration increased , reaching a plateau at 50 μM ( Fig 1C ) . As expected , uninfected cells did not exhibit an increase in fusion in response to the drug [24] , and the level of fusion present among HSV-infected cells treated with an equal concentration of DMSO matched that of untreated infected cells . Fusion of salubrinal-treated , KOS-infected cells reached ~50% ( Fig 1C ) , and this was even higher for strain 17-infected cells ( Fig 1B ) . A potential explanation for how salubrinal stimulates cell fusion was that its action disrupts protein complexes to allow the core fusion machinery ( gD , gH/gL , and gB ) to act independently from viral regulatory proteins ( Fig 1A ) . To rule out this possibility , we used viral mutants ΔUL16 and gEΔCT , where UL16 and the cytoplasmic tail of gE were deleted , to examine their effect on salubrinal-induced cell fusion [18 , 29] . Neither mutant was able to induce cell fusion with salubrinal treatment ( Fig 1D ) . Thus , it appears that accessory proteins work together with the core fusion machinery during salubrinal-induced fusion . Quantitation of syncytia can be accomplished by manually scoring nuclei ( Fig 1C and 1D ) ; however , this is time consuming and limits the number of samples that can be analyzed . Previously , an alternative method for quantitating cell fusion made use of a Coulter counter to measure the disappearance of single-nucleated cells as more cells fuse over time [30] . Since flow cytometry offers greater analytic power than a Coulter counter , we attempted to use it for rapid , quantitative measurements . Strain 17-infected Vero cells were treated with salubrinal or DMSO , and when extensive fusion was visually evident ( 12 hr post infection ) , the cells were exposed to trypsin for 5 minutes and subsequently harvested by vigorous pipetting in a solution of 2% paraformaldehyde to generate a homogenous suspension . The samples were then exposed to propidium iodide ( PI ) , which was able to stain all the nuclei since the cells were fixed . The suspensions were analyzed with a flow cytometer to measure particle numbers and sizes ( by forward light scattering , FSC-A ) , and the resulting dot plots ( S1 Fig ) were converted into histograms ( Fig 2A ) . In the DMSO control , one peak was observed representing the size distribution of intact , single cells present in the suspension ( Fig 2A ) . In contrast , a dramatically smaller-size population representing 75–80% of the PI-positive events was found for the salubrinal-treated sample ( Fig 2A ) . We hypothesized that these smaller particles were nuclei released from syncytia during sample preparation . To test whether the salubrinal-induced , PI-stained population behaved like free nuclei , a parallel culture of non-fused , HSV-infected cells were harvested and treated with 1% Triton X-100 to remove cell membranes; thereby releasing all the nuclei . These were pelleted , resuspended , and stained with PI prior to flow cytometry analysis . The detergent-released nuclei and most of the particles from salubrinal-induced syncytia were found to be of the same sizes , indicating that free nuclei indeed represent the majority of particles released from the sample of fused cells ( Fig 2B ) . To verify these results accurately reflect the number of nuclei contained in syncytia , duplicate cultures were infected with strain 17 and treated with salubrinal . At 12 hr post infection , one set of cells was fixed and manually analyzed , and the other was harvested , processed via flow cytometry , and the percentage of free nuclei was calculated . Both methods delivered the same results ( Fig 2C ) . To ascertain whether flow cytometry could be used to analyze fusion in another cell line , strain 17-infected BHK21 cells were treated with salubrinal . Massive cell fusion was apparent upon visual examination , and flow cytometry analysis revealed free nuclei levels of 75–80% , similar to what was seen with salubrinal-induced fusion in Vero cells ( Fig 2D ) . In contrast , transfected C10 cells ( used below ) were not amenable to flow cytometry because of clumping issues . With a rapid fusion assay in hand , we confirmed that strains 17 and F are more responsive to 50 μM salubrinal than KOS ( Fig 3A ) . Unfortunately , there are too many genetic differences to explain why KOS is somewhat limited in salubrinal-induced fusion [31] . For all three strains , we found that the extent of fusion increased with the multiplicity of infection ( MOI ) , suggesting that infected cells prefer to fuse with other infected cells ( Fig 3A ) . This is reminiscent of cells infected with gKsyn mutants , which prefer to fuse with other infected cells [30] in contrast to gBsyn mutants which have optimal fusion at lower MOIs . Beyond HSV-1 , cells infected with wild-type pseudorabies virus ( PRV , a porcine alphaherpesvirus ) were also stimulated to fuse with salubrinal to similar levels as the HSV-1 KOS strain , but this virus was not investigated any further . To determine the optimal timing of salubrinal treatment , two distinct time course experiments were conducted . In the first , cells infected with strain 17 at a high MOI received salubrinal at various times post infection , and the cultures were assayed for fusion at 12 hr post infection . This showed that salubrinal could be added as late as 4 hr with no reduction of fusion ( Fig 3B ) . When salubrinal was added at 6 hr , fusion levels dropped to 50% , and no fusion was observed when treatment began at 8 or 10 hr , since by that point , cells were rounding up due to extensive CPE and could no longer fuse . In a reciprocal experiment , salubrinal was added to cells immediately following strain 17 infection and then removed at various times . No fusion was observed when salubrinal was removed prior to 4 hr post infection ( Fig 3C ) ; however , total cell fusion reached 50% when salubrinal was removed at 6 hr , and maximal fusion was observed when salubrinal was removed after 8 hr . These data indicate that salubrinal exerts its effect only when the virus enters the late stage of infection , when the structural proteins of the virus begin to be synthesized [32] . Salubrinal treatment is well known to inhibit protein synthesis by increasing eIF2α phosphorylation levels in HSV-infected cells relative to a DMSO control , and this was confirmed in our studies ( Fig 4A ) . It has sometimes been assumed that salubrinal prolongs eIF2α phosphorylation by binding to GADD34 ( and the HSV-1 homolog , ICP34 . 5 ) , thereby preventing it from associating with PP1 . However , recent studies emphasize that salubrinal does not , in fact , bind directly to GADD34 [27 , 33] . In contrast , the drug guanabenz does act on GADD34 [27 , 34] , and to test whether it can stimulate fusion , KOS or strain 17-infected cells were incubated with various concentrations . No cell fusion was observed , even at the highest drug concentration tested ( 100 μM ) . Moreover , no increase in phosphorylation of eIF2α was observed ( Fig 4A ) , unlike what was found in previous studies of uninfected cells treated with guanabenz [34] . However , both salubrinal and guanabenz exhibited antiviral activity with titers being reduced by 1 log in KOS-infected cells ( Fig 4B ) . These results emphasize the importance of finding the host protein that binds salubrinal so that the mechanism by which it induces fusion can be understood . In an attempt to identify the target of salubrinal , a biotinylated chemical analogue named UTX-102 was synthesized ( S2A Fig ) . Importantly , this drug displayed similar activity to salubrinal in an established killing assay [35] with MDA-MB-231 breast cancer cells ( S2B Fig ) . These cells were treated with UTX-102 , cell lysate was harvested , and avidin agarose beads were used to pull-down protein complexes . Among the proteins identified by LC-MS/MS was protein tyrosine phosphatase 1B ( PTP1B ) , and immunoblot analysis verified that it was in the complex pulled down with UTX-102 ( S2C Fig ) . PTP1B is an ER-bound tyrosine phosphatase known to modulate a number of different signaling pathways , including events at the plasma membrane [36] . If inhibition of PTP1B in HSV-infected cells is what triggers cell fusion , then treatment with other known inhibitors of this enzyme should do the same . To test this , cells were infected with strain 17 and treated with two different allosteric inhibitors of PTP1B: TCS-401 or inhibitor XXII ( CAS 765317-72-4 ) [37 , 38] . Both drugs failed to stimulate fusion . Because the cellular functions of PTP1B are known to be regulated ( e . g . , by phosphorylation; [36] ) , we next considered the possibility that salubrinal might alter substrate specificity of this enzyme . If this were the case , then PTP1B inhibitors might block salubrinal-induced fusion . Accordingly , strain 17-infected cells were incubated with a combination of salubrinal and increasing amounts of each PTP1B inhibitor . In both cases , salubrinal-induced fusion was dramatically decreased with fusion being almost completely ablated at the highest concentrations of PTP1B inhibitors ( Fig 4C ) . While this suggests that PTP1B is critical for the fusion-inducing activity of salubrinal , it does not prove that this host factor is the target of salubrinal . Indeed , other results ( see below ) suggest that PTP1B is not the target . As inhibitor XXII more effectively limited fusion at lower dosages than the other drug , it was used in the subsequent experiments . Since all HSV-1 fusion events require the core fusion machinery ( gD , gH/gL , and gB ) , we next asked whether inhibitor XXII would block their activity . Transfected C10 cells , which overexpress the nectin-1 receptor for gD , fused efficiently when the four fusion proteins are co-expressed ( S3 Fig , top row ) , and manual counting revealed an average of 70 nuclei per syncytium ( Fig 4D , DMSO control ) . This did not change when inhibitor XXII was present ( Fig 4D and S3 Fig , bottom row ) . Transfected cells were also treated with salubrinal , and while there was slightly less fusion at 16 hrs post transfection ( Fig 4D ) , by 24 hr the level reached , but did not exceed , that of the control ( S3 Fig , middle row ) . Clearly , PTP1B and the target of salubrinal do not regulate the core fusion machinery in isolation but only when it is in a complex with accessory proteins ( Fig 1A ) . To get a glimpse at whether the effects of salubrinal and inhibitor XXII are limited to herpesviruses , cells infected with PIV5 , a paramyxovirus , were treated with the drugs . Neither salubrinal ( S4A Fig ) nor inhibitor XXII ( S4C Fig ) stimulated cell fusion . However , experiments with a fusogenic variant , rPIV5-NPΔ4v6 [39] , yielded different results . At 3 days post infection , multiple large syncytia were apparent in the DMSO controls , but to our surprise , salubrinal treatment actually reduced cell fusion ( S4B Fig ) . A reduction in the number and sizes of rPIV5-NPΔ4v6 syncytia was also seen with inhibitor XXII ( S4D Fig ) , which is in line with the ability of this inhibitor to block salubrinal-induced fusion of HSV-1 . Further investigation of paramyxoviruses is warranted , but we next examined the effects of these two drugs on HSV-1 Syn mutants . Syn variants arise from alterations to any of four different proteins ( gB , gK , UL20 , or UL24; Fig 1A ) , and these changes dysregulate the cell-to-cell spreading machinery in ways that are not understood , but are likely at distinct steps . We previously constructed examples of Syn mutants within the KOS strain [7 , 18] , and their responses to salubrinal and inhibitor XXII were measured . The gBsyn variant contained a single substitution , A855V , in the cytoplasmic tail , and at an MOI of 1 , infected cells exhibited 20% fusion at 12 hr post infection ( Fig 5A ) . Fusion levels rose dramatically with increasing concentrations of salubrinal , reaching a maximal 75% ( Fig 5A ) , which is much higher than the response seen with wild-type KOS ( Fig 3A ) . Hence , gBsyn and salubrinal are synergistic . To examine the impact of inhibitor XXII , we used a lower MOI and measured fusion at 24 hr post infection , which enabled gBsyn to reach 37% total fusion in the DMSO control ( Fig 5B ) . As gBsyn prefers to drive fusion with adjacent uninfected cells , the baseline fusion is enhanced at a lower MOI . Treatment with inhibitor XXII dramatically reduced fusion ( Fig 5B ) . These responses of gBsyn emulate what happens to wild-type HSV-1-infected cells when they are exposed to salubrinal and inhibitor XXII . Syncytial variants of gK cause even higher levels of fusion than those of gB , regardless of the MOI used during infection . Indeed , control cells infected with variants A40V or L118Q exhibited 80% total fusion ( Fig 5C and 5D ) . Because of these intrinsically high levels , we expected salubrinal to have no effect on the gKsyn phenotype; however , that was not the case . Instead , this drug greatly inhibited fusion , particularly in the case of L118Q ( Fig 5C ) . Thus , gKsyn mutants behave similarly to the fusogenic paramyxovirus variant described above . With regard to inhibitor XXII , it did produce the expected result and reduced gKsyn-mediated fusion in a dose dependent manner ( Fig 5D ) . Hence , both gBsyn and gKsyn seem to depend upon PTP1B for their Syn phenotype . Their very different responses to salubrinal is perhaps not surprising since their syncytial phenotypes differ in the requirement for tegument for UL21 [7] and perhaps other accessory proteins . The Syn variants of UL20 and UL24 have intrinsically weaker phenotypes , but their fusogenic activities were stimulated by salubrinal ( Fig 5E and S5 Fig , respectively ) . However , the maximal amount of fusion ( ~40% ) did not exceed the level observed when the wild-type KOS virus was exposed to salubrinal ( Fig 3A ) . Hence , the alterations that give rise to UL20syn ( substitution F222A ) and UL24syn ( G121A ) are not synergistic with salubrinal , and thus , the host factor targeted by this drug seems to work elsewhere on the regulatory machinery of the virus . Because the UL20syn and UL24syn variants are poorly fusogenic , we did not expect to be able to reliably measure any reductions in syncytia formation resulting from inhibitor XXII . Hence , we were not surprised to find that UL24syn was unaffected by this drug ( S5 Fig ) . However , to our considerable surprise , inhibitor XXII actually stimulated fusion for cells infected with UL20syn ( Fig 5F ) . From this , we predict that the change in UL20 causes a dramatic change to the configuration of the viral machinery such that the substrates of PTP1B are altered . All the experiments up to this point merely show that PTP1B is important for the Syn phenotype , whether induced by salubrinal or viral mutations; however , the overall goal of our experiments was to find host factors that are important for cell-to-cell spread . To examine the importance of PTP1B for this , we used wild-type virus at a low MOI and neutralizing antibodies to prevent cell-free spread . To visualize the sites of infection , the cells were fixed at various times and stained with antibodies against VP5 , the major capsid protein . Control experiments without inhibitor XXII ( Fig 6A and 6B ) confirmed that wild-type virus still forms large plaques via cell-to-cell spread in the presence of neutralizing antibodies , although their development is delayed . In contrast , the plaque sizes for strains KOS and 17 were dramatically decreased as the amount of inhibitor XXII increased ( Fig 6C and 6D ) . We also examined cell-to-cell spread in HaCaT cells , which are a keratinocyte-derived cell line that forms tight junctions and is relevant for HSV-1 infection [40] . With neutralizing antibodies preventing cell-free spread , inhibitor XXII once again dramatically reduced the average plaque size ( Fig 6G and 6H ) . A reduction in plaque size would also be expected if PTP1B was needed for infectious virion production . To address this possibility , Vero cells were infected at an MOI of 5 and treated with 30 μM inhibitor XXII , which is the minimal dose that gave the maximal effect in the cell-to-cell spread assay ( Fig 6C and 6D ) . The total amount of infectious virus produced was measured over time , and although there was a slight drop in titers at the later time points , the virus titers during the first 12–18 hours were not affected ( Fig 6E and 6F ) . Moreover , the amount of infectious virus released into the culture medium versus that remaining cell associated was unaffected by inhibitor XXII ( S6A Fig ) , and hence , PTP1B is not required for virion egress . The effect of adding inhibitor XXII prior to attachment and entry of the virus was also examined . In this case , cells were pretreated with DMSO or inhibitor XXII for one hour , and then dilutions of the virus were added for another hour , still in the presence of DMSO or drug . The culture medium was then replaced to remove the drug , and the cells were incubated to allow plaque formation . No reduction in the numbers of plaques was seen with 30 μM inhibitor XXII , and only a small effect was observed at 50 μM ( S6B Fig ) . Collectively , these results suggest that PTP1B activity is not required for binding , entry , replication , or egress . Rather , it appears that inhibitor XXII specifically affects the ability of the virus to move between cells . We further explored the drop in virus titer observed at later time points in the replication experiments . This effect was exaggerated for both strain 17 and KOS when 50 μM of XXII was used , even though the initial rates of infectious virion production were unaffected ( S6C Fig ) . We hypothesized that HSV-1 particles contain discrete binding sites for inhibitor XXII , and as these become saturated , the virus is inactivated . To test this idea , equal amounts of strain 17 were mixed with DMSO or inhibitor XXII , and the virions were incubated at 37°C . Samples were removed at various times and were serially diluted for plaque assays ( S6D Fig ) . Relative to the DMSO control , the 30 μM XXII treatment had no effect on virus titer for the first 12 hours and had only a small effect after that; however , at a concentration of 50 μM , the drug had a dramatic effect , and the virus was completely inactivated by 18 hrs . Although the mechanism of virus killing warrants further investigation , we sought ways to eliminate PTP1B activity without the use of inhibitors . Fortunately for this investigation , PTP1B-knockout mice have been made and found to develop into adults [41 , 42] . Moreover , immortalized mouse embryo fibroblasts ( MEFs ) have been derived from these mice ( PTP1B-/- ) , along with a reconstituted line ( PTP1B+ ) in which the coding sequence has been reinserted with a retroviral vector [43] . These two lines were obtained , and immunoblots with an antibody specific for PTP1B confirmed their identities ( S7 Fig ) . If PTP1B is not required for HSV replication , then there should be no defect in viral replication on PTP1B-/- cells , and that is what we found ( Fig 7A ) . Also , PTP1B+ cells infected with strain 17 exhibited robust cell fusion when treated with salubrinal , although the development of syncytia took a few hours longer than with Vero cells , and this was blocked by inhibitor XXII ( Fig 7B ) . As expected , the knockout cells were not very responsive to salubrinal , and although a small induction of fusion was observed , this was eliminated with inhibitor XXII ( Fig 7B ) . Hence , another tyrosine phosphatase similar to PTP1B ( e . g . , TC-PTP ) [44] might be able to compensate . To test the ability of the two MEF lines to support cell-to-cell spread , they were infected at low MOI with strain 17 or KOS , and neutralizing antibodies were added to the medium to block cell-free spread . At 42 hr post infection , the average plaque size on PTP1B-/- cells was reduced by more than 60% for both viruses compared to their sizes on reconstituted PTP1B+ cells ( Fig 7C ) . A somewhat greater reduction in plaque size was seen when strain 17-infected PTP1B+ cells were treated with inhibitor XXII , but there was no further decrease in plaque size in the parallel culture of PTP1B-/- cells ( Fig 7D and 7E ) . These experiments , along with the inhibitor XXII studies in Vero and HaCaT cells , confirm that PTP1B is required for cell-to-cell spread of HSV-1 . To try and gain mechanistic insight , KOS-infected HaCaT cells were treated with DMSO or XXII and gE , E-Cadherin , and VP5 were examined with confocal microscopy . We found that gE ( S8A Fig ) and E-Cadherin ( S8B Fig ) localization was unaffected by PTP1B inhibition , and gE trafficking to the cell junctions ( where it is known to accumulate ) was not impeded . Additionally , VP5 accumulation at cellular junctions did not appear to be increased in cells treated with inhibitor XXII ( S8B Fig ) . Ultimately , it appears that PTP1B inhibition does not block cell-to-cell spread by mislocalizing gE or altering capsid trafficking . The small increase in cell fusion observed when infected knockout cells were treated with salubrinal ( Fig 7B ) suggested that the cellular target of this drug might still be present . If PTP1B is not the target , then eIF2α should be phosphorylated to the same extent in response to salubrinal in both MEF lines . For this experiment uninfected cells were used , and these were treated with DMSO , salubrinal , or thapsigargin , which is a well-known inducer of ER stress [45] . The basal level of eIF2α phosphorylation was higher for the knockout cell line ( Fig 7F ) , but this was expected since PTP1B is needed to fully shut down the eIF2α kinase PERK [46] . Thapsigargin-induced stress pushed both cell lines into states with equally high levels of phosphorylation , and these levels were matched by salubrinal treatment ( Fig 7F ) . Thus , PTP1B seems unlikely to be the target of salubrinal .
HSV-1 capsids are wrapped with host-derived membranes in the cytoplasm to produce mature virions within vesicles [48] , and these subsequently fuse with the plasma membrane to release their contents for cell-free spread . As shown here , PTP1B is not required for any of the steps in that pathway , but instead is involved only in cell-to-cell spread . The prevailing model for cell-to-cell spread merely requires transport of virion-containing vesicles to lateral cell junctions , where fusion with the plasma membrane releases HSV-1 into intercellular spaces so that the adjacent cells can be infected [20] . This simple model does not specify how many virions can be delivered into the adjacent cells or whether those cells have a mechanism for excluding the passage of subsequent virions , as is the case for cell-free infections , where the “door closes” two hours after the initial infection [49] . Also , this model does not specify any role for viral proteins on the plasma membrane within cell junctions . In support of the current model , mature virions are observed in spaces between cells [50] . Also , receptor nectin-1 resides in cell junctions , where the virus can bind ( via gD ) to gain entry into adjacent cells [51] . However , because HSV-1 downregulates nectin-1 in infected cells , this receptor is only present on the adjoining uninfected cell [52] . Consequently , virions released into lateral junctions ( as opposed to viral proteins on the plasma membrane ) are unlikely to be the drivers of fusion for Syn mutants or in salubrinal treatments because they cannot fuse with the cell from which they emerged [52] . Further supporting the current cell-to-cell spread model , the tail of gE , which forms a heterodimer with gI ( Fig 1A ) and is likely exposed on the cytoplasmic faces of virion-containing vesicles [53] , appears to be required for transport to lateral junctions [54] . That is , mutants lacking the tail of gE have been reported to release their virions at apical membranes rather than at cell junctions [50 , 54] . There are many reasons to suspect that cell-to-cell spread is more complicated than merely transporting virions to cell junctions . For example , the list of viral proteins that seem to participate in the spreading mechanism is quite long and includes ( just for example ) four tegument proteins that directly interact with the tail of gE , namely UL11 , UL16 , VP22 , and UL51 [18 , 19 , 55] . The current model for cell-to-cell spread does not take into account this complexity . Unfortunately , mutational studies of these and all the other viral proteins that have been implicated in the spreading mechanism ( including gE/gI ) are difficult to interpret because these proteins play multiple roles in the virus replication cycle ( capsid envelopment , for one example ) . Moreover , these proteins exist in poorly-defined complexes with one another , and those can fall apart when all or a portion of one subunit is missing [56 , 57] . Also , the viral proteins that mediate cell-to-cell spread may assemble into unique complexes with unexpected properties . For instance , pseudorabies virus ( PRV , another neurotropic herpesvirus ) does not require gD for cell-to-cell spread , even though this glycoprotein is essential for infectivity . That is , a null mutant propagated in complementing cells to temporarily provide gD , can infect cell cultures or pigs , where it spreads cell-to-cell efficiently , but all of the progeny viruses lack gD and are noninfectious for cell-free spread [58 , 59] . Thus , at least for PRV , the machinery used for cell-to-cell spread is not identical to that used for virus entry . There is strong evidence that the membranes used to wrap capsids in the cytoplasm are derived from endosomes , and therefore , it is likely that these carry all the viral membrane proteins , which would have been at least transiently present on the cell surface [60] . Moreover , studies too numerous to cite have shown that many viral proteins indeed accumulate on the plasma membrane within cell junctions . For example , detailed studies have shown that gE/gI and gB are redistributed to cell junctions at late times after infection [61] , and here we have shown that gE is present , even when PTP1B is inactive . Of particular relevance ( explained below ) , gN/gM complexes are also found at cell junctions [62] . In addition to directing its own proteins to cell junctions , HSV-1 induces a remodeling of cellular proteins in adherens junctions . For example , the connections provided by nectin-1 subunits , which dimerize to link cells , are lost when gD downregulates the subunits in infected cells , thereby freeing up their partners in adjacent cells to allow infection [51 , 52 , 63] . Also , cellular membrane proteins TGN46 and carboxypeptidase D have been reported to accumulate at cell junctions [61] . A role for modified-cell junctions has been long imagined; in particular with regard to gE/gI [20 , 64] , but it remains unclear what is accomplished for the mechanism of cell-to-cell spread . It is not obvious why viral fusion machinery would be needed at cell junctions to support the release of virions contained within secretory vesicles . Outside of the herpesvirus family , measles virus induces the formation of an intercellular pore , which facilitates the transfer of cytoplasm from infected to uninfected cells and potentially provides a pathway for the direct transfer of virions between cells [2 , 65] . While there is no evidence that HSV capsids pass through specialized pores during cell-to-cell spread , it cannot be ruled out , and it is intriguing that neurons infected with PRV form small pores between cells in a manner that is dependent upon the viral fusogen , gB [66] . Viral membrane proteins may also be needed to create contacts between cells where they do not normally exist . This is perhaps easiest to understand with sensory neurons , which HSV-1 enters to establish latent infections and leaves after reactivation , employing cell-to-cell spread in both directions . However , sensory neurons respond to mechanical forces and temperature ( for example ) but do not normally establish synapses with epithelial cells [67] . Thus , it seems likely that viral proteins on the surface of infected cells are needed to establish connections to enable the passage of HSV-1 , perhaps in a manner analogous to the unique connections created by HIV for cell-to-cell spread [68] . In this regard , it is interesting that PTP1B plays a role in cell adhesion and migration [69 , 70] , which may permit infected cells to make local changes in position to establish contacts with nearby uninfected cells . However , cell-to-cell spread occurs in cells that already do have cell junctions , and it is likely that unique , virus-specific connections are needed in this situation , too . Of course , the presence of viral fusion machinery at cell junctions requires that it be tightly regulated; otherwise , syncytia would form . While the regulatory mechanism is virtually unknown , our experiments confirm that salubrinal treatments disrupt it , but they go further in showing that tyrosine phosphorylation plays a critical role . In addition , the differential responses of the four types of Syn mutants to salubrinal and PTP1B inhibitor XXII ( illustrated in S9 Fig ) make it clear that they each dysregulate the viral fusion machinery in distinct ways , all of which deserve further investigation . PTP1B has been shown to be anchored in the endoplasmic reticulum with its catalytic domain positioned in the cytosol where it modulates a wide variety of cellular functions , including the unfolded protein response [36 , 71 , 72] . We cannot rule out the possibility that PTP1B plays a role in the transport , docking , or fusion of vesicles at cell junctions so that their virion cargoes can be released . We did not observe an obvious accumulation of capsids at cell junctions by confocal microscopy in the absence of PTP1B activity , but that might be expected if transport to cell junctions was impaired . However , the importance of this enzyme in influencing the properties of Syn mutants and salubrinal treatments strongly suggests that it plays a role in modulating viral machinery on the plasma membrane within cell junctions . This is not surprising because PTP1B is well known to regulate cellular receptors on the plasma membrane [73] and is recruited to regions of cell-cell contact to bind N-cadherin and stabilize its association with beta-catenin at adherens junctions [74 , 75] . In view of our findings , it is intriguing that gM , which also regulates the viral fusion machinery at cell junctions along with its disulfide-bonded partner , gN [62] , has very recently been found to interact with “extended synaptotagmin 1” [76] . This host protein helps connect the ER to the plasma membrane [77] , and it might play a role in positioning PTP1B at the proper sites for its regulatory activity in the cell-to-cell spreading mechanism . There are three critical areas that need to be pursued to further elucidate how PTP1B participates in the spreading mechanism . One is the identification of the proteins that contain the critical tyrosine substrates . Fortunately a PTP1B trapping mutant is available that can bind but not dephosphorylate [78] . Insertion of the coding sequence for this mutant into the wild-type HSV-1 genome would allow it to be expressed in all infected cells , and if a suitable tag is fused to the construct , then the bound substrates can be recovered and identified by mass spectrometry with known cellular targets of PTP1B serving to validate the approach [79] . Viral substrates would be particularly interesting , as would any novel cellular substrates , all of which would have to be confirmed as being important for cell-to-cell spread and syncytia formation . Another critical question is whether the absence of PTP1B blocks cell-to-cell spread of wild-type HSV-1 in animals . Mouse lines are already available in which the gene for PTP1B is knocked out in all tissues [41] , and other lines are available that have the gene flanked by LoxP sites so that expression can be knocked out in specific tissues [80] , with neuronal and epithelial cells being of particular interest . There are well-established methods for investigating HSV infection in vivo [81 , 82] , and it should be straightforward to determine whether and how the absence of PTP1B affects cell-to-cell spread in vivo . It is possible that this enzyme would only be needed in infected cells , but perhaps it is needed in target cells , too . If PTP1B is in fact needed for cell-to-cell spread in mice , then existing inhibitors ( e . g . , MSI-1436 , see above ) might prove to be beneficial for treatment of HSV-1 reactivation disease in immunocompromised patients . One of the biggest missing pieces to this story is the target of salubrinal . Our initial candidate was PTP1B , but when this host protein is absent , phosphorylation of eIF2α still occurs normally in response to salubrinal . Our data suggest that the target for salubrinal is contained within a protein complex that includes PTP1B , and it is possible that it will turn out to be a tyrosine kinase . However , if the target of salubrinal is not a kinase , then there must be three host factors involved: PTP1B , a tyrosine kinase , and the target of salubrinal . Lastly , all herpesviruses have mechanisms for cell-to-cell spread , none of which are understood . It is possible that the host factors implicated here in the spreading mechanism of HSV-1 may be relevant for other viruses . Indeed , our data suggest that PTP1B and the target of salubrinal may even play a role in paramyxoviruses .
Vero ( African green monkey kidney ) cells ( a gift from Richard Courtney , Penn State University ) were grown in Dulbecco’s modified Eagle’s medium ( DMEM; Gibco ) supplemented with 5% bovine calf serum ( BCS; HyClone ) , 5% fetal bovine serum ( FBS; HyClone ) , and penicillin-streptomycin ( pen/strep; Gibco ) . HaCaT ( human ) cells ( a gift from Craig Meyers , Penn State University ) , C10 ( mouse ) cells ( a gift from Gary Cohen , University of Pennsylvania ) , and BHK-21 ( Syrian golden hamster ) cells ( a gift from Nicholas Buckovich , Penn State University ) were grown in DMEM supplemented with 10% FBS and pen/strep . MDA-MB-231 ( human ) breast cancer cells ( ATCC , Manassas , VA ) were grown in DMEM with 10% FBS and pen/strep . PTP1B-/- mouse embryonic fibroblasts ( MEFs ) ( a gift from Benjamin Neel , NYU ) were cultured in DMEM supplemented with 10% FBS and pen/strep , and the recombinant PTP1B+ MEFs also received hygromycin B ( 200 μg/mL; ThermoFisher ) to select for the retroviral vector [43] . All cells were maintained in a humidified incubator at 37°C in 5% CO2 . Infected cells were cultured in DMEM supplemented with 1% FBS . The KOS and F strains of HSV-1 were produced by transfecting Vero cells with BACs containing the viral genomes [83] and harvesting the virus produced by transfected cells ( transfection stock ) . Virus stocks were generated by infecting Vero cells with the transfection stocks at an MOI of 0 . 01 and harvesting at 48 hours post by freezing and thawing the cultures 3 times to maximize the release of virions . A virus stock of strain 17 was a gift from Moriah Szpara ( Pennsylvania State University ) . HSV-1 mutants ΔUL16 , gEΔCT , gB . A855V , gK . A40V , gK . L118Q , UL20 . F222A , and UL24 . G121A were previously constructed [7 , 18 , 29] in the KOS genome by means of BAC recombineering [84] . A virus stock of the PRV Becker strain was a gift from Lynn Enquist ( Princeton University ) . Titers of all HSV variants and PRV were measured by plaque assays on Vero cells . Virus stocks of wild-type PIV5 and mutant rPIV5-NPΔ4 were produced as previously described [39] . For experiments with HSV-1 and PRV , Vero cells or BHK-21 cells were seeded into 6-well plates at a density of 4 . 2 x 105 cells/well and cultured for 2 days . Once confluent , the cells were infected at the indicated MOIs , which ranged from 0 . 1 to 3 depending on the experiment . During the 1-hour infection period , cells were incubated at 37°C and rocked every 15 minutes . After infection , the cells were rinse once with DMEM and incubated at 37°C with media containing DMSO ( 5 μl/ml ) ( Sigma-Aldrich ) , salubrinal ( Sigma-Aldrich ) , or inhibitor XXII ( CAS 765317-72-4; Santa Cruz ) . Experiments were conducted with equal volumes of DMSO in each sample , reaching a final concentration of 5 μl/ml of DMSO . Incubation time ranged from 12–24 hours depending on the experiment . Because fusion was slower to develop in PTP1B-/- MEFs , measurements were taken at 24 hours post infection . In the paramyxovirus experiments , Vero cells were plated at 1 . 2x106 cells per well in 6-well dish one day prior to infection with rPIV5 NPΔ4v6 at MOI of 0 . 1 or rPIV5 at MOI of 1 . Viruses were allowed to adsorb for one hour at 37°C , and then the cells were washed twice with 1X PBS , followed by replacement with high-glucose DMEM supplemented with 2% FBS , pen/strep , and either DMSO , 30 μM PTP1B inhibitor XXII , or 50 μM salubrinal . Cells were returned to the incubator and syncytia formation monitored . At various times post infection , monolayers were visualized using a Nikon Eclipse TS100-F microscope and photographed using a Nikon DS-Fi1 digital camera . To manually quantify the amount of cell fusion , cells were fixed for 10 minutes in 4% paraformaldehyde ( PFA; Sigma-Aldrich ) , rinsed two times in PBS , permeabilized with 0 . 1% Triton X-100 for 10 minutes , and subsequently blocked for 30 minutes in 2% BSA/PBS ( Sigma-Aldrich ) . To visualize the cell perimeters , the samples were incubated with a mouse monoclonal antibody specific for ZO-1 ( 1A12; ThermoFisher ) at a dilution of 1:500 for one hour . After rinsing 2X with PBS , the samples were stained with a fluorescently-conjugated secondary antibody ( g-α-m Alexa568; Life Technologies ) at a dilution of 1:1000 for one hour , and nuclei were stained with DAPI ( Molecular Probes ) for 5 minutes . Images were captured with an Olympus IX73 inverted microscope . Fusion was calculated by manually counting the number of nuclei present in single cells versus the number of nuclei contained within syncytia with the aid of Olympus cellSens software . At least 1000 nuclei were scored per image for each experimental replicate . To quantify cell fusion in a more rapid and unbiased manner , a novel flow cytometry method was developed . For this , cultures in 6-well plates containing syncytia were gently rinsed with standard buffer , and then 400 μl of trypsin in standard buffer ( Sigma-Aldrich ) was added to each well . The plate was rocked every minute until the majority of cells were lifting off the plate , and then the cells were fixed by adding 600 μl of ice-cold 4% PFA/PBS solution ( EMS; Corning ) to each well for a total volume of 1ml . Importantly , all buffers used were Ca2+/Mg2+-free to minimize cell clumping . The solution was pipetted up and down around the entire well 10–20 times to break up syncytia and achieve a homogenous suspension . The samples were transferred to pre-chilled flow cytometry tubes on ice , and the wells were rinsed with 200 μl of FACS buffer ( 2% BSA , 3 mM EDTA in Ca2+/Mg2+-free PBS ) , which was added to the respective sample tubes . Nuclei were stained by adding 200 μl of propidium iodide staining solution ( 100 μg/ml in FACS buffer; ThermoFisher ) to each sample . Flow cytometry tubes were vortexed for 3 seconds , placed on ice , and covered with foil to limit light exposure . The samples were passed through a BD LSRFortessa cell analyzer within 1 hour of harvesting . 50 , 000 PI-positive events were collected per sample , and the data were analyzed with FlowJo software . Briefly , PI-positive events were gated for FSA ( forward scatter ) and by SSC ( side scatter ) using a logarithmic scale . Events were then categorized as free nuclei or intact cells having single nuclei based on size and granularity ( S1 Fig ) . The percentage of free nuclei was used to calculate total cell fusion within a sample . Vero cells were seeded in 6-well plates at a density of 4 . 2 x 105 cells/well , and after reaching confluency ( ~1 . 2 x 106 cells/well ) , infections were initiated with KOS or strain 17 at an MOI of 5 in a volume of 500 μl for 1 hour at 37°C . To inactivate virions remaining on the cell surface , the cultures were briefly rinsed with a low pH buffer ( 135mM NaCl , 10mM KCl , 40mM citric acid , pH 3 . 0 ) and then rinsed two times with DMEM . The cells were then incubated at 37°C in 1 ml/well of DMEM ( +1% FBS ) containing DMSO or the specified amounts of salubrinal , guanabenz ( Sigma-Aldrich ) , or inhibitor XXII . At 6 , 12 , 18 , 24 , and 30 hours post infection , a cell scraper was used to harvest the total cells and media for each sample . Not all time points were used for every experiment and differences are noted in the accompanying figure legends . Each sample was processed through 3 freeze/thaw cycles prior to serial dilution and titration by plaque assay . For MEFs , infections were started when cell density reached 1 . 3 x 106 cells/well and a citric acid wash was not used . For the virus egress assay , infected cells and media were harvested and titered separately . Vero cells , HaCaT cells , or MEFs were seeded onto glass coverslips in 6-well plates , allowed to reach confluency , and infected with ~100 PFU/well of KOS or strain 17 . After 1 hour , the cells were rinsed twice with DMEM and incubated in infection media ( DMEM + 1% FBS ) containing 5 mg/ml pooled human IgG ( Equitech-Bio , Inc ) . This concentration of IgG was previously determined to neutralize all but 2 virions per 1x106 PFU [7] . At 42 hpi , the cells were fixed for 10 minutes in 4% paraformaldehyde ( PFA; Sigma-Aldrich ) , rinsed twice in PBS , permeabilized with 0 . 1% Triton X-100 for 10 minutes , and blocked for 30 minutes in 2% BSA/PBS ( Sigma-Aldrich ) . The samples were stained with a rabbit antibody against VP5 ( the major capsid protein ) at a dilution of 1:1000 for 1 hour , rinsed 3 times with PBS , and stained with an Alexa 568 fluorescent secondary antibody ( Life Technologies ) for 1 hour at a dilution of 1:1000 . Finally , samples were stained with DAPI for 5 minutes , rinsed 3X with PBS , and the coverslips were mounted on slides using Aqua Poly/Mount ( Polysciences , Inc . ) . Fluorescent images of plaques were captured using an Olympus IX73 inverted microscope , and their sizes were measured using Olympus cellSens software . A biotinylated derivative of salubrinal , UTX-102 ( S2A Fig ) , was synthesized and is similar to ones previously described [85] . MDA-MB-231 cells were incubated with 20 μM UTX-102 for 5 hours , and cell pellets were lysed in buffer ( 100 mM Tris-HCl pH 7 . 5 , 150 mM NaCl , 1% Triton X-100 , and 5% glycerol ) supplemented with phosphatase ( Sigma-Aldrich , St . Louis , MO , USA ) and protease ( Thermo-Fisher Scientific , Waltham , MA , USA ) inhibitors . Avidin-biotin pull-downs were carried out with a Pierce Monomeric Avidin kit ( Thermo-Fisher Scientific , Waltham , MA , USA ) . Briefly , samples were incubated in the column to which the monomeric avidin was immobilized , and these were washed with PBS until absorbance at A280 came to baseline , indicating that most non-specific binding proteins were removed . Bound proteins were eluted , concentrated , desalted , denatured , and alkylated according to standard protocols . The samples were digested with trypsin overnight ( 37°C ) , and peptides were desalted with a Silica C18 MacroSpin Column , reconstituted in 5% ACN , 0 . 1% formic acid , and injected onto a ODS-100V C18 column ( Tosoh Bioscience ) . The peptides were eluted with a linear gradient , and the effluent was electro-sprayed into a LTQ Orbitrap mass spectrometer . Blanks were run prior to the sample run to ensure that no significant background signals existed from solvents or the columns . Database search and data analysis were performed via Thermo-Fisher Scientific Proteome Discoverer™ software ( v1 . 3 ) against Uniprot protein database ( Uniprot_072413_HUMAN . fasta ) . To detect PTP1B , cells were lysed in standard RIPA buffer containing protease inhibitors . Proteins were separated in a SDS-10% polyacrylamide gel and electro-transferred to Immobilon-P or nitrocellulose membranes , which were incubated first with a primary antibody against PTP1B ( rabbit polyclonal , #5311S; Cell Signaling ) and then with a secondary antibody conjugated with horseradish peroxidase . Protein levels were assayed with a SuperSignal West Femto Maximum Sensitivity substrate ( Thermo Scientific ) or Pierce West Pico PLUS reagents . To assess the phosphorylation state of eIF2α , Vero cells or MEFs were treated with DMSO , salubrinal , guanabenz , or thapsigargin as indicated , and the cells were lysed on ice for 10 minutes in protein extraction buffer ( 10 mM Tris hydroxymethyl aminomethane , pH 6 . 8 , 150 mM NaCl , 1 mM EDTA , and 0 . 5% Igelpal; Sigma ) containing protease ( Sigma #P8340 ) and phosphatase ( PhosStop; Roche ) inhibitor cocktails at recommended concentrations . Debris was removed with a 21K x g spin at 4°C for 10 minutes , and the protein concentration of each supernatant was measured by the Pierce BCA assay . Samples containing 20 μg were electrophoresed in an SDS-12% polyacrylamide gel , and transferred to a sheet of nitrocellulose , which was subsequently blocked with 5% BSA in TBS-T ( Tris-buffered saline , pH 7 . 6 , 0 . 1% Tween 20 ) . The total amounts of eIF2α were measured with a 1:1000 dilution of a primary mouse monoclonal antibody ( #2103S; Cell Signaling Technology ) and a secondary HRP-conjugated goat-anti-mouse antibody . The levels of phosphorylated eIF2α were measured with a 1:1000 dilution of a primary rabbit monoclonal antibody ( #3398S; Cell Signaling Technology ) and a secondary HRP-conjugated goat-anti-rabbit antibody . Chemiluminescence signals were detected with Pierce West Pico PLUS reagents using BioRad ChemiDocMP and quantified with ImageLab 6 . 0 Software . HaCaT cells were seeded onto coverslips and infected with the KOS strain at an MOI of 0 . 1 . Cells were incubated in infection media containing DMSO or 30 μM drug XXII for 18 hours at 37°C , fixed in 4% paraformaldehyde ( PFA; Sigma-Aldrich ) for 10 minutes , and rinsed twice in PBS . Then , they were permeabilized with 0 . 1% Triton X-100 for 15 minutes , blocked for 1 hour in 2% BSA/PBS ( Sigma-Aldrich ) , and exposed for 1 hour to primary antibodies against gE ( mouse monoclonal 3114 , kindly provided by David Johnson , Oregon Health & Science University ) at a 1:4000 dilution , E-cadherin ( mouse monoclonal , #610818; BD Biosciences ) at a 1:300 dilution , or VP5 ( rabbit polyclonal ) at a 1:1000 dilution . After rinsing 3 times with PBS , the cells were stained with secondary fluorescent goat antibodies ( Alexa 488-anti-mouse or Alexa 568-anti-rabbit; Life Technologies ) and DAPI ( Molecular Probes ) . The coverslips were mounted onto slides with Aqua Poly/Mount ( Polysciences , Inc . ) , and Z-stack images were obtained with a Nikon C2+ confocal microscope and processed using Nikon Elements software . All statistical analyses were performed using Prism ( GraphPad , v4 ) . A two-tailed Student’s T-test was used to determine statistical significance and the number of replicates performed for each experiment is listed in the figure legends . FlowJo was used to analyze all flow cytometry data , Nikon Elements was used to process confocal images , and Olympus cellSens was used for manual counting of nuclei . | It is estimated that 67% of the global population is infected with herpes simplex virus type 1 ( HSV-1 ) . This virus resides in sensory neurons in a quiescent state but periodically reactivates , producing virus particles that travel down the axon to infect epithelial cells of the skin , where it can be transmitted to additional people . To avoid neutralizing antibodies , herpesviruses have evolved mechanisms for moving directly from one cell to another through their sites of intimate contact; however , the mechanism of cell-to-cell spread is poorly understood . Studies of HSV-1 mutants have implicated numerous viral proteins , but the necessary cellular factors are unknown except for the one that the virus uses to enter cells . Our experiments have identified a cellular enzyme ( PTP1B , a tyrosine phosphatase ) that is dispensable for the production of infectious virions but is critically important for the cell-to-cell spreading mechanism . Promising drugs targeting PTP1B have already been tested in early clinical trials for possible treatment of obesity and type-2 diabetes , and thus , our study may have immediate utility for attenuating HSV-1 reactivation disease in immunocompromised patients . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
]
| [
"flow",
"cytometry",
"cell",
"physiology",
"phosphorylation",
"vero",
"cells",
"biological",
"cultures",
"microbiology",
"viral",
"structure",
"membrane",
"proteins",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"proteins",
"cell",
"lines",
"viral",
"replication",
"cell",
"membranes",
"virions",
"spectrophotometry",
"cytophotometry",
"biochemistry",
"cell",
"biology",
"post-translational",
"modification",
"virology",
"biology",
"and",
"life",
"sciences",
"spectrum",
"analysis",
"techniques",
"cell",
"fusion"
]
| 2018 | The HSV-1 mechanisms of cell-to-cell spread and fusion are critically dependent on host PTP1B |
Tolerance to high levels of ethanol is an ecologically and industrially relevant phenotype of microbes , but the molecular mechanisms underlying this complex trait remain largely unknown . Here , we use long-term experimental evolution of isogenic yeast populations of different initial ploidy to study adaptation to increasing levels of ethanol . Whole-genome sequencing of more than 30 evolved populations and over 100 adapted clones isolated throughout this two-year evolution experiment revealed how a complex interplay of de novo single nucleotide mutations , copy number variation , ploidy changes , mutator phenotypes , and clonal interference led to a significant increase in ethanol tolerance . Although the specific mutations differ between different evolved lineages , application of a novel computational pipeline , PheNetic , revealed that many mutations target functional modules involved in stress response , cell cycle regulation , DNA repair and respiration . Measuring the fitness effects of selected mutations introduced in non-evolved ethanol-sensitive cells revealed several adaptive mutations that had previously not been implicated in ethanol tolerance , including mutations in PRT1 , VPS70 and MEX67 . Interestingly , variation in VPS70 was recently identified as a QTL for ethanol tolerance in an industrial bio-ethanol strain . Taken together , our results show how , in contrast to adaptation to some other stresses , adaptation to a continuous complex and severe stress involves interplay of different evolutionary mechanisms . In addition , our study reveals functional modules involved in ethanol resistance and identifies several mutations that could help to improve the ethanol tolerance of industrial yeasts .
The ability to survive and proliferate in high levels of ethanol is an ecologically important and industrially relevant trait of yeast cells . The ethanol produced by yeast cells slows down growth of competing microbes , but at higher concentrations , it causes stress for the yeast cells themselves . Different yeast strains show significant differences in their ability to grow in the presence of ethanol , with the more ethanol-tolerant ones likely having a fitness advantage over non-tolerant strains [1–3] . Moreover , ethanol tolerance is a key trait of industrial yeasts that often encounter very high ethanol concentrations , for example during beer and wine making and industrial bio-ethanol production . Because of its industrial importance , there is great interest in fully understanding the genetic underpinnings of ethanol tolerance in microbes . Different experimental approaches have been used , including screening of ( deletion ) mutants for increased ethanol tolerance , transcriptome analysis of ethanol stressed cells and QTL analyses aimed at identifying mutations that cause differences in ethanol tolerance between different yeast strains [4–10] . Together , these studies have linked multiple different genetic loci to ethanol tolerance and identified hundreds of genes , involved in a multitude of cellular processes [11–14] . While it becomes increasingly clear that ethanol is a complex stress that acts on several different processes including increasing fluidity and permeability of cellular membranes , changing activity and solubility of membrane-bound and cytosolic proteins and interfering with the proton motive force ( for review , see [15–17] , the exact molecular mechanisms and genetic architecture underlying ethanol tolerance are still largely unknown . Experimental evolution to study adaptation to increased ethanol levels could provide more insight into the molecular mechanisms underlying ethanol tolerance since such experiments could reveal different mutational paths that make a sensitive strain more tolerant . Only a handful of studies have looked at adaptation to ethanol in originally non-ethanol tolerant microbes exposed to gradually increasing levels of ethanol [6 , 18–22] . These have mostly focused on the physiological adaptations found in the evolved cells and have not performed an extensive analysis of the mechanisms and genetic changes underlying this adaptation . Hence , a comprehensive analysis of the type and number of mutations a non-ethanol tolerant strain can ( or needs to ) acquire to become more ethanol tolerant is still lacking . Experimental evolution has proven to be a valuable tool to investigate the different mechanisms and pathways important for cells to adapt to specific selective conditions . Seminal papers have increased our understanding of the molecular basis of adaptation to specific stresses , such as heat stress , nutrient limitation and antibiotic treatment [23–28] . Recent advances in DNA sequencing technologies allow affordable and fast sequencing of complete genomes of clones and populations . While sequencing clones yields information on individual lineages within the experiment , population data provides information on the heterogeneity of adaptation . Additionally , sequencing samples isolated at different time points during the evolution experiment makes it possible to capture evolution in action . This has provided valuable information on the rate and types of mutations underlying adaptation , the genetic basis of ‘novel’ phenotypes and the existence of parallel pathways to establish comparable phenotypic outcomes [29–32] . For example , a common strategy observed in populations evolving under nutrient limitation is the amplification of genetic regions encoding transporters responsible for the uptake of the limiting nutrient [24 , 33] . Other studies using multiple replicate populations have discovered a high degree of parallelism in the adaptive solutions found by different populations . However , in other cases , the evolutionary pathways can be more complex . For example , clonal interference , the competition between lineages carrying different beneficial mutations , is a commonly observed phenomenon in evolution of asexually propagating populations that can increase complexity of mutational dynamics as well as impede the spread of beneficial mutations in a population . Clonal interference is typically expected to be prevalent in large populations that are adapting to a complex stress , where multiple adaptive mutations can occur [32 , 34 , 35] . To unravel the molecular mechanisms of adaptation to a specific condition , most studies have used isogenic replicate populations , with all cells having the same initial genome size . Genome size can significantly change during evolution; with both small-scale changes ( chromosomal deletions and amplifications ) and large-scale changes ( increase or decrease in ploidy ) . Moreover , ploidy shifts have been reported in the evolutionary history of many organisms , including Saccharomyces cerevisiae and as a response to selective pressure [36–38] . Conversely , genome size has also been reported to affect evolution rate: polyploidy has been suggested to increase adaptability [39 , 40] . Multiplying the amount of DNA increases the genetic material available for evolution to tinker with and can alter gene expression [41–43] . These polyploid genomes can be unstable , resulting in loss of chromosomes and thus aneuploid cells [44–46] . Although studies have looked at adaptation of lineages of different ploidy , none have followed the mutational dynamics in these evolving populations over time in detail [36 , 47–49] . In this study , we use experimental evolution to dissect the adaptive mechanisms underlying ethanol tolerance in yeast . Six isogenic Saccharomyces cerevisiae populations of different ploidy ( haploid , diploid and tetraploid ) were asexually propagated in a turbidostat over a two-year period , with ethanol levels gradually increasing during the experiment . This step-wise increase in exogenous ethanol levels resulted in a constant selective pressure for our cells . Whole-genome sequencing of evolved populations and isolated , ethanol-tolerant clones at different times during the experiment allowed us to paint a detailed picture of the mutational dynamics in the different populations . Several common themes in the type of adaptations and evolutionary mechanisms emerge , with all lines showing extensive clonal interference as well as copy number variations . Additionally , the haploid and tetraploid lines showed rapid convergence towards a diploid state . Despite these common themes , we find multiple lineage-specific adaptations , with little overlap in the mutated genes between the different populations . By applying a novel computational pipeline to identify affected pathways , we were able to reveal overlap between the functional modules affected in the different adapted populations , with both novel and previously established pathways and genes contributing to ethanol tolerance . Importantly , introduction of specific mutated alleles present in adapted populations into the ancestral strain significantly increased its ethanol tolerance , demonstrating the adaptive nature of these mutations and the potential of using experimental evolution to unravel and improve a complex phenotype .
To study the mutational dynamics underlying increased ethanol tolerance , six prototrophic , isogenic S . cerevisiae strains of different ploidy ( two haploid , two diploid and two tetraploid lines—with each line of the same initial ploidy started from the same preculture ( VK111 , VK145 and VK202 respectively , see also S4 Table ) were subjected to increasing levels of ethanol . Specifically , ethanol levels were gradually increased from 6% ( v/v ) to 12% over a two-year period in a continuous turbidostat with glucose ( 4% ( w/v ) ) as a carbon source . Samples were taken at regular time intervals and subjected to whole-genome sequencing analysis . Chloramphenical was added to the medium of our long-term evolution experiment to prevent bacterial growth . Chloramphenicol does not affect yeast growth ( S1 Fig ) . In addition to sequencing each of the 6 evolving populations , we also sequenced the genomes of three clones isolated from each of the population samples , resulting in a total of 34 population samples and 102 clonal samples that were sequenced . Fig 1 depicts the experimental set-up as well as the time points ( number of generations ) for which whole-genome sequencing was performed . We determined the relative fitness of the evolved populations , isolated at 40 and 200 generations , and generally observed increases in fitness in high ( 9% v/v ) EtOH ( Fig 2 & Materials and Methods ) . In general , clones taken from the same population at the same time point had similar fitness; with some notable exceptions suggesting considerable heterogeneity within populations ( S2 Fig ) . These fitness measurements were performed in 9% ethanol because the ancestral reference strains did not grow at higher ethanol levels . It is important to note that the actual increase in fitness of our evolved strains in higher ethanol levels ( above 9% ( v/v ) ) is much larger than what can be appreciated from Fig 2 . Spotting of our evolved lineages on agar plates with increasing levels of ethanol indicates that these adapted strains can grow on concentrations up to 11% ( v/v ) EtOH , with the haploid control ancestral strain showing almost no growth under these conditions ( i . e . an infinite increase in fitness compared to the ancestral haploid strain ) ( S3 Fig ) . Ancestral diploid and tetraploid strains show growth on 11% EtOH , although they are still outperformed by their evolved clones under these conditions . Propidium iodide staining and flow cytometry analysis of evolved populations showed that diploid cells appeared relatively quickly in the originally haploid populations ( Fig 3; PI staining profiles of ancestral strains can be found in S4 Fig; ploidy of sequenced clones can be found in S1 File ) . This was observed independently in both reactors started from the same isogenic haploid population . It should be noted that around generation 40 in reactor 1 , some diploids are already present in the originally haploid population . By generation 60 , haploid cells took over the population again , possibly by acquiring a mutation that made them more fit than these new diploids . In both reactors , diploid cells have taken over the entire population by 130 generations . These diploid variants are still of mating type α , the same as the initial haploid ancestral strain , indicating that they did not arise through mating type switching and subsequent mating . Our results indicate that the diploids in our turbidostat are likely the result of failed cytokinesis . The relatively rapid evolutionary sweep of the new diploid variants points to a significant selective advantage . Competition experiments confirmed that the fitness of a diploid cell is indeed significantly higher than that of an isogenic haploid strain in 9% EtOH ( p = 0 . 0063; unpaired t-test ) , whereas there is no significant fitness difference in the absence of ethanol ( p = 0 . 469; unpaired t-test ) ( S5 Fig ) . Interestingly , our tetraploid starting populations also converged to a diploid state relatively fast during the evolution experiment . The parallel diploidization observed in two reactors with haploid ancestral cells and two reactors with tetraploid ancestral cells might suggest that becoming diploid is one of the main and/or one of the more easily accessible routes leading to increased ethanol tolerance . To identify the mutational pathways during adaptation to ethanol , we performed whole-genome sequencing of populations of evolving cells throughout the two-year experiment . Each population sample was sequenced to 500 fold coverage on average , and this for multiple time points ( 34 samples for all reactors combined , see Fig 1 ) . In addition to this whole-population sequencing strategy , we also sequenced for each time point the genomes of three adapted clones to 80 fold coverage on average , resulting in 102 clonal samples sequenced in total . Details on the sequencing and analysis pipelines can be found in Materials and Methods . After 2 years , yielding around 200 generations , evolved clones ( excluding clones from reactor 2 and 6 , which acquired a mutator phenotype , see below ) contained on average 23 SNPs compared to the ancestral strain ( S1 File ) . This number is higher than what would be expected based on measured rates of spontaneous mutations [50] , and could reflect an increased mutation rate under the stressful conditions imposed by the high ethanol levels in our set-up . Across all reactors and clones sequenced , we identified a total of 8932 different sites mutated . The largest fraction are SNPs ( 6424 out of 8932; 72% ) ; Indels are found mostly in non-coding regions ( 1830 out of 2508; 73% ) , whereas SNPs are mostly found inside genes ( 4971 out of 6424; 77% ) . Most of these coding SNPs are non-synonymous ( 3672 out of 4971; 74% ) . A full list of all mutations ( SNPs and Indels ) in all sequenced clones for the different reactors and time points can be found in the S1 File . In two reactors ( reactor 2 and reactor 6 ) , we noticed a marked increase in the number of mutations found in individual clones ( see Fig 4 , S1 and S2 Files ) . Specifically , clones from reactor 2 contain a significantly higher number of indels ( about 929 on average per genome for clones isolated after 200 generations , compared to 21 on average per genome for clones isolated after 200 generations in other reactors ) , whereas clones from reactor 6 display an increase in the number of SNPs ( about 1765 per genome for clones isolated after 200 generations , compared to an average of 31 per clone for other reactors ) . This high number of mutations points to a so-called “mutator” phenotype typically present in cells with lower DNA replication fidelity and/or DNA repair . Interestingly , such mutators are frequently observed during evolution experiments , probably because mutators are more likely to acquire ( combinations of ) adaptive mutations faster [23 , 51–54] . Interestingly , the increased mutation rate in reactor 2 is first observed around generation 70–80 , coinciding with the appearance of a 21 bp insertion in the MSH2 gene– a key player in DNA mismatch repair . Mutations in MSH2 have been previously reported to confer a mutator phenotype [55 , 56] , and mutations in MutS , the ortholog of MSH2 in E . coli , were identified in a long term evolution experiment and also result in a mutator phenotype [23] . Indeed , deletion of this gene , as well as re-creating this insertion in an otherwise wild type background drastically increases mutation rate ( S6 Fig ) . The MSH2 mutation eventually reaches a frequency of 100% at the end of the evolution experiment ( 200 generations ) . We currently do not know the exact cause of the elevated mutation rate observed in reactor 6 . Apart from the convergence towards diploidy ( see above ) , we also detected extensive copy number variation ( CNVs ) in our evolved clones , comprising both duplicated and deleted chromosomal regions ( see Fig 5 and S3 File ) . Other studies have observed similar copy number variation during adaptation to stress , including heat stress and specific nutrient limitations [24 , 28 , 33] . Interestingly , acquisition of an extra copy of chromosome III appeared to be a common feature for most of our evolved clones ( S3 File ) . Some ( parts of ) other chromosomes , including chromosome XII and chromosome IV , are also duplicated in several of our evolved clones . These frequent occurrences indicate that these specific aneuploidies could be adaptive under ethanol conditions , although the exact mechanistic basis remains to be elucidated and further experiments are needed to validate the ( potential ) adaptive role of these observed aneuploidies . While the sequencing of clones yielded valuable information on individual lineages within the evolving populations; sequencing of evolving populations yielded more information on the complex mutational paths and dynamics between sub-populations within each evolving population . A list of SNPs and Indels identified in the evolved populations , together with their allele frequency in the population at each sampling point , can be found in S4 File . We observe distinct pattern of mutations appearing and disappearing over time in each of the six reactors . Some of these mutations remain in the population , eventually reaching high levels or even complete fixation ( i . e . presence in 100% of all cells in the population ) . Other mutations only persist for a short time , until lineages carrying these mutations are outcompeted by others , so-called clonal interference . In total , we identified 1637 mutations across all populations and time points . 117 of these mutations are no longer present in the final time points sequenced , and 101 mutations drop more than 10% in frequency after reaching their maximum frequency , indicative of clonal interference . Interestingly , we also identified some overlap between the mutations found in different independently evolving populations ( i . e . in different reactors ) . Specifically , we find 20 genes that are mutated twice in different generations and populations , 3 genes mutated 3 times , 2 genes mutated 4 times , 2 genes mutated 5 times and 1 gene 6 times ( see also S1 Table ) . This significantly differs from what would be expected by chance ( see Materials and Methods and S1 Table for p-values , exact binomial test ) . Repeatedly hit genes are , amongst others , involved in stress response , cell cycle and heme biosynthesis . The higher number of sequenced samples from reactors 1 and 2 allowed us to further analyze these population sequences and group mutations based on correlations in the changes in their respective frequencies ( based on the pipeline developed by [32]; see also Materials and Methods ) . This yields a more detailed picture of the different co-evolving sub-populations present in these reactors , which is depicted in the Muller diagrams of Fig 6 and S7 Fig , see also S2 Table for haplotype frequencies . In both reactors , selective sweeps mostly consists of groups of mutations that move through the population together . While these reactors were inoculated with the same strain , the type and dynamics of mutations observed during adaptation appear very different . However , both evolving populations of reactor 1 and reactor 2 are characterized by strong clonal interference . In reactor 1 , 4 different subpopulations are present around generation 90 , each carrying different mutations . By generation 130 , these lineages have been outcompeted by another lineage that almost completely dominates the population by 200 generations ( S7 Fig ) . In reactor 2 , a lineage carrying a mutation in PDE2 ( encoding a high-affinity cAMP phosphodiesterase ) is driven to extinction by a subpopulation carrying indels in ASG1 and MSH2 . ASG1 is a transcriptional regulator involved in the stress response and has been found mutated in evolved populations from different reactors ( including reactor 1 , see also Fig 6 , S7 Fig and S4 File ) . Another example of clonal interference is observed in the later generations of this reactor: a subpopulation carrying a mutation in CDC27 ( encoding a subunit of the anaphase promoting complex/cyclosome ) is driven to extinction by a subpopulation carrying mutations in BNI1 ( important for nucleation of actin filaments ) and PET123 ( encoding a mitochondrial ribosomal protein ) . The results discussed above revealed extensive variability in the type and number of mutations present in each evolving population . While this could suggest the presence of several , different mutational pathways ( and thus lack of parallel evolution ) , mutations in different genes might affect identical or similar pathways , implying that the physiological adaptation to high ethanol might in fact be more similar than what is immediately apparent from the individual mutations . It should also be noted that some of the mutations identified in our evolved lineages might represent adaptations to the device used , or other aspects of the selection regime , rather than ethanol per se . To gain insight into the affected biological pathways and investigate the possible similarities in adaptation to increased ethanol , different computational approaches were used ( see also S1 Text ) . First , a term-enrichment analysis was performed ( for enriched clusters , see Table 1 and S5 File ) . For this analyses , we excluded mutations obtained from the populations with a mutator phenotype ( reactor 2 and 6 ) because of their low signal to noise ratio . These enrichment methods have been used as one of the standard functional analysis tools and gave us a first insight into potential adaptive pathways present in our evolved lineages . In a second step , we used a sub-network-based selection method [58 , 59] ( see Materials and Methods ) first developed for E . coli expression data . Here , we adapted and extended this method to select the subnetwork from the global yeast interaction network that best connected the mutated genes in the most parsimonious way . This method also identifies the intermediary genes involved in signaling mechanisms , which are not necessarily mutated in our evolved lineages but mediate the cellular response . We excluded mutations obtained from the populations with a mutator phenotype ( reactor 2 and 6 ) because of their low signal to noise ratio . This analysis identifies genes frequently mutated in the different populations ( DSK2 , ASG1 ) , as well as genes that are closely connected on the interaction graph ( HEM3 , HEM12 , … ) . This latter set reflects parallelism at the pathway level in the different reactors . 25% of genes from inside our identified networks have been previously linked to ethanol tolerance ( broadly defined as fermentation/growth capacity under conditions with EtOH , as reported in literature ) , whereas such as a connection was found for only 9% of genes outside of the enriched networks . From Fig 7 , it is clear that sometimes multiple mutations occurring in the same pathway originated in the same reactor rather than across different reactors . This could indicate not only a level of parallelism between the different reactors but also between the sub-populations in the same reactor . To confirm this , we also applied the sub-network selection method on the mutations obtained for each of the populations separately . This confirms that within single populations the same pathways seem to be affected as those that are consistently affected between the populations ( including cell cycle , DNA repair and protoporphyrinogen metabolism , see also below ) . All resulting sub-networks as well as the enriched genes ( including p-values ) can be found in S6 File , Fig 7 and S8–S13 Figs . To increase mechanistic insight , we performed several additional analyses , such as interactome analyses of genes containing non-synonymous mutations in our different population samples as well as analyses of the effect of these mutations on protein functional domains ( see also S1 Text , S3 Table and S7 File for more details ) . Together , these analyses helped us identify the different pathways that were affected in our evolved lineages . Pathways identified as affected across different reactors by these different analyses are discussed in more detail in the next paragraphs . The high number of mutations ( both SNPs and Indels ) precluded an exhaustive analysis of all mutations present in our tolerant clones and populations and their effect on ethanol tolerance . One commonly used approach to investigate putative beneficial mutations is backcrossing of evolved clones to their ancestor , which does not contain any mutation . However , since each of the evolved populations proved unable to form any spores , this strategy was not accessible to us . Hence , we focused on SNPs reaching high frequencies in our 200 generation population samples . In total , nine SNPs were selected for further study ( Table 2 ) . Several of the mutated genes belong to or are linked to the processes affected across and/or within specific reactors ( heme metabolism , protein transport and cell cycle , see also Table 2 ) . These nine SNPs were subsequently introduced into the ancestral haploid strain . The effect of these mutations was assessed by high-throughput competition experiments [64] , in 0 , 4 , 6 and 8% ( v/v ) EtOH conditions , with glucose as a carbon source ( Fig 8 , S14 and S15 Figs and S6 Table ) . Several mutants show a clear increase in fitness , with the fitness effect often depending on the concentration of ethanol in the medium . Most mutants show slightly increased fitness in medium without added ethanol , with further increases in relative fitness with increasing exogeneous ethanol concentrations . Mutants in PRT1 and MEX67 are less fit than the WT in non-ethanol conditions , but show increased fitness in higher ethanol levels . MEX67 is a poly ( A ) RNA binding protein involved in nuclear mRNA export . PRT1 encodes the eIF3b subunit of the eukaryotic translation initiation factor 3 ( eIF3 ) . The increased ethanol tolerance associated with these mutations hints at translation processes as targets of ethanol . Interestingly , a recent study in E . coli showed that ethanol negatively impacts transcription as well as translation [6] . However , if and how mutations in MEX67 and PRT1 might mitigate this effect and contribute to ethanol tolerance remains unknown . Of all mutations introduced into the ancestral strain , vps70C595A provided the largest fitness increase . VPS70 is putatively involved in sorting of vacuolar carboxypeptidase Y to the vacuole [65] . A mutation in VPS70 has been recently identified by members of our team as a determinant of ethanol tolerance in a Brazilian bioethanol strain [66] . Interestingly , the VPS70 mutation in our evolved ethanol-tolerant lineages alters the same amino acid as the mutation present in the industrial ethanol tolerant strain ( Goovaerts A . and Thevelein J . M . , personal communication ) . The fact that one of the mutations identified in our evolved lineages was also found in an industrially used bio-ethanol strain underscores the potential of our approach to find biologically relevant mutations for increased ethanol tolerance . Other members of the VPS family have been previously implicated in ethanol tolerance as well , but also here the exact molecular mechanism through which they could increase ethanol tolerance is still unclear [10 , 12 , 67] . To our knowledge , none of the other genes investigated in this study have been previously implicated in ethanol tolerance nor have these mutations been found in natural or industrial yeasts so far . This implies that these mutations could be prime candidates to improve the ethanol tolerance and production of existing industrial yeasts [68] .
Many fundamental questions on the dynamics and genetics of adaptation to a complex and severe stress such as ethanol are still unanswered . Which processes contribute to ethanol tolerance ? Which type and number of mutations are needed and/or sufficient , and how do they penetrate and fix within populations ? To what extent are these mutations and the pathways they affect predictable ? To address these questions , we performed high-coverage whole-genome sequencing of clones and populations isolated throughout a two-year evolution experiment and characterized their adaptation to increasing ethanol levels . This allowed us to paint a detailed picture of the mutational dynamics in our evolving lineages . Our study demonstrates how many different evolutionary mechanisms all come together to provide adaptation to a severe and complex stress . Specifically , we find that adaptation to high ethanol levels involves changes in ploidy , copy number variation and the appearance of mutator phenotypes , with evolving populations showing strong clonal interference . Although these mechanisms have been observed in other studies ( see , for example , [24 , 27 , 32 , 69–71] , different mechanisms are often observed and/or studied separately . More importantly , in traditional evolution studies , populations are exposed to a fixed concentration of antibiotic or limiting nutrient . Under these conditions , selection pressure is reduced or even eliminated when cells become resistant . In our experimental set-up , ethanol levels were stepwise increased over time so that cells were evolving under constant selection . High concentrations of ethanol also actively kill non-tolerant cells , so that cells not adapted to the higher ethanol concentrations die and/or are washed out of the turbidostat . Taken together , our study combines aspects of traditional evolution studies with principles used in morbidostat experiments [27] . This set-up is likely a better representation of what happens when a population gradually penetrates and adapts to a new niche . Under such severe and continuous stress conditions ( near-morbidostats ) , many cells fail to survive . This also implies that the traditional way of measuring evolutionary time by counting the number of generations may become problematic as cell death and mutations contributing to survival instead of replication become increasingly important . This phenomenon can also result in extremely quick sweeps , because a mutation cannot only penetrate a population because it increases the replication rate , but also because it protects against death . Moreover , as cell replication slows down , the number of mutations that is not associated with DNA replication but rather with DNA damage and repair likely increases . It seems plausible that these mechanisms contribute to the very quick genetic sweeps that we observe in some reactors , most notably the quick sweep of diploidized cells in reactors 1 and 2 . This could also explain why the number of mutations that we observe in our two-year experiment might seem high for only 200 generations when compared to experiments where cells are growing at higher rates and where 200 generations are often reached within only one or two months of evolution . While it is difficult to think of a better universal measure of evolutionary time , it is important to realize that in some cases , a combination of physical time and generations may be a more appropriate way to assess evolution . We find extensive variation in genome size in our evolved cells; with aneuploidies in a large number of our evolved clones as well as convergence toward a diploid state in our initially haploid and tetraploid populations . Becoming diploid thus appears to be a frequently used strategy by cells of different ploidy which effectively increases ethanol tolerance . Convergence to diploidy has been observed in other laboratory evolution experiments [25 , 36 , 72] . In contrast to these studies , we could demonstrate a clear fitness advantage of our diploid strains in ethanol . Although further research is needed to fully understand this fitness advantage , changes in gene expression and/or cell size could be important factors contributing to the increased fitness of diploid cells [41 , 73 , 74] . Unfortunately , this convergent evolution prevented us from performing a detailed analysis on the difference in adaptive strategies employed by haploids , diploids and tetraploids to increase ethanol tolerance . Aneuploidy and copy-number variation are increasingly recognized as common themes in rapid adaptation [24 , 28 , 69 , 70] . It is believed that changes in the copy number of chromosomes or chromosomal fragments provide a relatively easily accessible way to change expression levels of specific key genes [75 , 76] , and these CNVs can provide large , usually condition-specific , fitness effects [77] . However , such large-scale changes in the genome likely also have some unwanted , detrimental side effects , such as imbalance between gene products and genome destabilization [78–80] . Evolving populations are believed to gradually replace adaptive CNVs with more specific mutations that show fewer pleiotropic effects [28] . Notably , several of our evolved clones , isolated from different reactors , carry an extra copy of chromosome III and/or ChrXII; pointing towards a potential adaptive benefit of this specific aneuploidy . Interestingly , we find that clones isolated at later time points have a specific , smaller region of chromosome XII amplified ( see also S3 File ) ; indicating a more refined solution . GO enrichment and network analyses of the repeatedly amplified region of ChrXII ( position 657500–818000 ) hints at cell wall formation as one of the key processes affected by these amplifications . Previous studies have indeed shown that cell wall stability is a key factor involved in ethanol tolerance [18] . Apart from diploidization of our evolving lineages , another example of parallelism at the phenotypic level is the appearance of a mutator phenotype in two of our six evolving populations . The sweep of the MSH2 mutation in reactor 2 is likely caused by a so-called hitchhiking event , with the high mutation rates in the MSH2 mutant leading to the appearance of one or several beneficial mutations that drive the selective sweep . Parallelism at the genotypic level is less clear: we find few mutations and mutated genes shared between the different evolved lineages . Interestingly , nine ( ACE2 , POL3 , PUF4 , GFA1 , UTH1 , JID1 , RIM15 , ATG11 and VPS74 ) out of the 28 genes ( 32% ) that were found hit in multiple reactors have been previously implicated in ethanol tolerance . Applying different types of network and enrichment analyses revealed functional modules affected in several of the adapted populations . These pathways include response to stress , intracellular signal transduction , cell cycle and pathways related to membrane composition and organization ( such as isoprenoid metabolism , glycerophospholipid catabolism and fatty-acyl-coA metabolism ) . For some of these pathways , further work is needed to clarify their exact involvement in ethanol tolerance . To investigate the phenotypic effect of mutations present in our evolved lineages , we performed high-throughput fitness measurements in different ethanol concentrations . While our evolved lineages contained multiple mutations , single mutations reaching high frequency in the evolved populations could already significantly increase ethanol tolerance when introduced into the ancestral , non-evolved haploid strain . Moreover , several of these single mutations selected for further study also showed a ( modest ) fitness benefit in conditions with no ethanol , with greater increases as ethanol levels rise . The number of mutations identified in this study was too large to investigate the fitness effect of each individual mutation . However , our strategy to select mutations that reached fixation in the evolved populations clearly proved successful , identifying as many as 4 mutations ( out of 9 tested ) that confer a fitness advantage in high ethanol environments . Mutations in genes linked to processes identified as affected across our different reactors—such as protein transport ( VPS70 ) –indeed increased fitness in EtOH . This underscores the potential of PheNetic , a sub-network based selection method for identifying adaptive mutations . Five of the mutations tested did not significantly increase fitness , although they reached high frequency in our adapted populations . This is indicative of hitch-hiking with other , beneficial mutations; or possible epistatic interactions with other mutations . Why then would not all feral yeasts show high ethanol tolerance , if it appears so easy to attain ? Firstly , it seems plausible that not all yeasts are confronted with selection for high ethanol tolerance . Furthermore , it is important to note that we have not tested the fitness of the mutants under many different conditions that mimic the natural habitats of yeasts . It seems likely that some of the mutations identified in this study would result in lower fitness in other environments [81 , 82] . Moreover , we have not investigated the effect of combined mutations . While it is possible that combining different mutations could increase ethanol tolerance even further , it also seems likely that some mutations and/or specific combinations of mutations could lead to reduced fitness in different low and/or high ethanol environments . Indeed , while our clones have increased fitness in EtOH , we observe that fitness of several of our evolved clones ( containing multiple other mutations apart from the ones investigated in this study ) decreased in medium without exogenous EtOH ( see S16 Fig ) . These results are indicative of antagonistic pleiotropy: the specific mutations present in our evolved clones increase fitness in one condition ( high ethanol , which was selected for ) , whereas they reduce fitness in other environments . Ethanol resistance is an important trait for the survival of feral yeasts in nature because the ethanol produced inhibits growth of competing microorganisms , while it serves as a carbon source in later stages of growth , when all fermentable sugars are depleted ( the so-called make-accumulate-consume strategy [83–85] ) . Our results suggest that adaptation to high ethanol is complex and can be reached through different mutational pathways . Apart from yielding insight into the evolutionary mechanisms leading to such complex and ecologically important phenotypes , our study is also of considerable industrial importance . Several of the mutations identified in this study may be useful to increase the ethanol tolerance of industrial strains used for the production of alcoholic beverages or biofuels .
Starting strains for the evolution experiment are all derived from the haploid prototrophic S288c strain FY5 [86] . To prevent clumping of cells during the evolution experiment , the flocculation genes FLO1 , FLO10 and FLO11 were deleted in this strain using deletion cassettes based on pUG6 , conferring resistance to G-418 disulfate [87] . Markers were removed through the Cre/LoxP technique using pSH65 [88] . Mating type switching of this strain was then performed , using plasmid pSB283 , to create isogenic diploid and tetraploid strains . Fluorescent versions of strains ( YECitrine or mCherry tagged ) were constructed by integrating fluorescent markers at an intergenic , neutral region of chromosome II . A full list of strains used , with their complete genotype , can be found in S4 Table . Primers used for strain construction and verification can be found in S5 Table . Populations were founded in 400 mL ethanol containing media . Media contained 10 g/L yeast extract , 20g/L bactopeptone , 4% ( w/v ) glucose , 0 . 001% ( v/v ) Rhodorsil , Antifoam Silicone 46R , chloramphenicol ( 50 μg/mL ) and increasing concentrations of ethanol . Populations were maintained at an average population size of 1010 cells . After 25 generations , the level of EtOH in the media was increased each time ( starting at 6% ( v/v ) and reaching 12% at 200 generations ) . Turbidostat cultures were maintained using Sixfors reactors ( Infors ) at 30°C , pH was kept constant at 5 . 0 with continuous mixing at 250 rpm in aerobic conditions . At regular times , a population sample was obtained from each of the cultures for further analyses and stored in glycerol at -80°C . For DNA extraction purposes , a population cell pellet was also frozen down at -80°C . Fitness for all evolved strains was determined in rich medium ( YP , 2% ( w/v ) glucose ) with 9% ( v/v ) ethanol , by competing strains against a YECitrine labeled ancestral strain . Cultures were pre-grown in YPD 6% ethanol . After 12h , wells of a 96 deep well plate were inoculated with equal numbers of labeled reference and unlabeled strains ( ~ 106 cells of each ) and allowed to grow for around 10 generations . Outer wells only contained medium and acted as a buffer to prevent ethanol evaporation . Additionally , plates were closed with an adhesive seal and plastic lid , and parafilm was used to prevent ethanol evaporation . Cultures were regularly transferred to new medium to prevent nutrient depletion . The ratio of the two competitors was quantified at the initial and final time points by flow cytometry . Data analysis was done in FlowJo version 10 . Measurements were corrected for the small percentage of labeled , non-fluorescent cells that occurred even when the reference strain was cultured separately as well as for the cost of YFP expression in the labeled reference strain . For each fitness measurement , three independent replicates were performed . The selective advantage , s , of each strain was calculated as s = ( ln ( Uf/Rf ) -ln ( Ui/Ri ) ) /T where U and R are the numbers of unlabeled and reference strain respectively , the subscripts refer to final and initial populations and T is the number of generations that reference cells have proliferated during the competition . The fitness of the unlabeled WT strain was designated 1 , fitness of the evolved strains as 1+s . DNA content of evolved populations and evolved clones was determined by staining cells with propidiumiodide ( PI ) and analyzing 50 000 cells by flow cytometry on a BD Influx . The ancestral haploid and diploid strains used in the evolution experiment were used for calibration . For evolved populations , genomic DNA was directly extracted from pellets that were frozen at the time of sample taking . Evolved clones were selected from the different population samples by streaking glycerol stocks from the corresponding population samples on YPD plates . Swabs from each population were subsequently grown in YPD 6% EtOH and dilutions were plated on YPD plates with different ethanol concentrations ( ranging from 8% to 10%; with a 0 . 5% stepwise increase in ethanol concentrations ) . From these plates , ethanol tolerant clones were selected and genomic DNA of these clones was extracted . Genomic DNA was prepared using the Qiagen genomic tip kit . Final DNA concentrations were measured using Qubit . Paired-end sequencing libraries with a mean insert size of 500 bp were prepared and libraries were run on an Illumina HiSeq2000 ( EMBL GeneCore facility , Heidelberg ) . Average sequencing coverage for clone and population sample is 80X and 500X respectively . The p-values probabilities of genes being hit by a mutation a specific number of times were calculated using the binomial distribution . The number of draws was set to the total number of mutations found inside coding regions ( i . e . 817 ) ; the number of successes was set to the number of times a specific gene was hit by a mutation ( 2 to 6 ) ; the probability of success was set to the specific ratio of the length of each gene hit multiple times ( in nucleotides ) over the entire coding content of the S . cerevisiae genome ( 9080922 nucleotides ) . Haplotype reconstruction was based on the approach described by [32] . Prior to the actual reconstruction , variants identified in the sequencing data were subject to further processing by excluding variants that were multi-allelic , variants that exhibited mixed zygosity in the isolated clones ( i . e . homozygous in one clone and heterozygous in another clone ) and variants that didn't reach the frequency of 0 . 2 at any point during the experiment . Haplotypes ( or 'mutational cohorts' ) present in the remaining variants were next reconstructed using a Matlab script ( kindly provided by the Desai lab ) on a local machine running Matlab 2013a . Briefly , variants present in our dataset were clustered into haplotypes based on the Euclidean distance between their frequencies at specific time-points of the experiment . Afterward , frequencies of individual variants assigned to a specific haplotype were averaged to obtain the frequency of the haplotype itself . Frequencies of the identified haplotypes at specific time points ( see S2 Table ) were then used as source data to draw their approximate Muller diagram representations with Inkscape 0 . 48 . 4 . Functionally meaningful terms enriched in the list of genes hit in our evolution experiment were identified using DAVID Tools [57 , 92] . An enrichment score cutoff value of 1 . 3 was used , equivalent to a p-value of 0 . 05 for term enrichment , as recommended by the authors [57] . Intergenic mutations were discarded for the analysis . The interaction network used as input for PheNetic was composed of interactomics data obtained from KEGG [93] for metabolic interactions , String for protein-protein interactions [94] and Yeastract for protein-DNA interactions [95] . The total interaction network contains 6592 genes and 135266 interactions . This interaction network was converted to a probabilistic network using the distribution of the out-degrees of the terminal nodes of the network edges . By doing so , edges connecting nodes with a low out-degree will receive a high probability while edges connecting nodes with a high out-degree receive a low probability . Using lists of mutated genes as input , PheNetic will now infer that sub-network of the probabilistic network that best connects the mutated genes in the list over the probabilistic network . As the probabilistic network penalizes hub nodes , the inferred sub-network will therefore preferentially connect the mutated genes through the least connected parts of the network . This results in selecting the most specific parts of the network that can be associated with the mutated genes . PheNetic was used to connect the mutated genes over the interaction network [58] with the following parameters: 100-best paths with a maximum path length of 4 were sampled between the different mutations in combination with a search tree cutoff of 0 . 01 . As the size of the selected sub-network by PheNetic is dependent on both a cost parameter and the number of mutated genes in the input , different costs were used for the sub-network inference from different sizes of mutated gene lists . For the sub-network inference between all the mutated genes from the non-mutator reactors a cost of 0 . 25 was used , for the non-mutator reactors ( 1 , 3 , 4 , 5 ) a cost of 0 . 05 was used as they all have a similar amount of mutated genes , and for the mutator reactors ( 2 and 6 ) a cost of 0 . 5 was used . For more details , see S1 Text . The resulting networks were visualized using Cytoscape and a functional enrichment using the biological process terms of Gene Ontology in combination with the annotation of SGD of the sub-networks was performed using the Bingo plugin , version 2 . 44 [96] . The resulting enrichment results are listed in S6 File . Selected mutations identified from the whole-genome sequencing data were introduced into the ancestral genetic background using the following protocol . First , a selectable marker conferring resistance to hygromycine was introduced near the genomic location of interest through homologous recombination . Part of this locus was then amplified together with the selectable marker using a forward primer containing the desired mutation . The resulting PCR product was then transformed into the ancestral strain; and presence of the mutation was verified by Sanger sequencing . Primers used can be found in S5 Table . YECitrine- or mCherry-tagged site-directed mutant strains were competed with the parental mCherry or YECitrine tagged strains , respectively , as described [64] . In brief , saturated cultures of mutant and parental strains were mixed in equivalent volumes and inoculated onto 150 μl of YNB-low fluorescent medium in 96-well microtiter plates ( Corning 3585 ) . Micro-cultures grew without shaking and were serial-diluted every 24 hrs for approximately 28 generations ( 7 days ) in a fully automated robotic system ( Tecan Freedom EVO200 ) that integrates a plate carrousel ( Liconic STX110 ) , a plate reader ( Tecan Infinite M1000 ) , a 96-channel pipetting head , an orbital plate shaker , and a robotic manipulator arm . The equipment was maintained in an environmental room at constant temperature ( 30°C ) and relative humidity ( 70% ) . Fluorescence signal ( mCherry: Ex 587 nm/5 nm and Em 610 nm/5 nm; YECitrine: Ex 514 nm/5 nm and Em 529 nm/5 nm ) and absorbance at 600 nm were monitored every hour during the entire experiment . The YECitrine- or mCherry-tagged parental strains were competed to each other for normalization and monitored individually to determine background fluorescence signal . Fluorescence and absorbance output data was analyzed in Matlab as described [64] to obtain an average selection coefficient , smut , with its S . E . M . from three experimental replicates . All sequencing data are available from the NCBI Sequence Read Archive: Bioproject , accession number PRJNA292495 . | Organisms can evolve resistance to specific stress factors , which allows them to thrive in environments where non-adapted organisms fail to grow . However , the molecular mechanisms that underlie adaptation to complex stress factors that interfere with basic cellular processes are poorly understood . In this study , we reveal how yeast populations adapt to high ethanol concentrations , an ecologically and industrially relevant stress that is still poorly understood . We exposed six independent populations of genetically identical yeast cells to gradually increasing ethanol levels , and we monitored the changes in their DNA sequence over a two-year period . Together with novel computational analyses , we could identify the mutational dynamics and molecular mechanisms underlying increased ethanol resistance . Our results show how adaptation to high ethanol is complex and can be reached through different mutational pathways . Together , our study offers a detailed picture of how populations adapt to a complex continuous stress and identifies several mutations that increase ethanol resistance , which opens new routes to obtain superior biofuel yeast strains . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| []
| 2015 | Adaptation to High Ethanol Reveals Complex Evolutionary Pathways |
Staphylococcus aureus is an opportunistic pathogen that produces many virulence factors . Two major families of which are the staphylococcal superantigens ( SAgs ) and the Staphylococcal Superantigen-Like ( SSL ) exoproteins . The former are immunomodulatory toxins that induce a Vβ-specific activation of T cells , while the latter are immune evasion molecules that interfere with a wide range of innate immune defences . The superantigenic properties of Staphylococcal enterotoxin-like X ( SElX ) have recently been established . We now reveal that SElX also possesses functional characteristics of the SSLs . A region of SElX displays high homology to the sialyl-lactosamine ( sLacNac ) -specific binding site present in a sub-family of SSLs . By analysing the interaction of SElX with sLacNac-containing glycans we show that SElX has an equivalent specificity and host cell binding range to the SSLs . Mutation of key amino acids in this conserved region affects the ability of SElX to bind to cells of myeloid origin and significantly reduces its ability to protect S . aureus from destruction in a whole blood killing ( WBK ) assay . Like the SSLs , SElX is up-regulated early during infection and is under the control of the S . aureus exotoxin expression ( Sae ) two component gene regulatory system . Additionally , the structure of SElX in complex with the sLacNac-containing tetrasaccharide sialyl Lewis X ( sLeX ) reveals that SElX is a unique single-domain SAg . In summary , SElX is an ‘SSL-like’ SAg .
Staphylococcus aureus is a serious human pathogen responsible for a large proportion of hospital acquired infections and of additional major concern , an increasing cause of community-associated antibiotic resistant infections [1 , 2] . Predominantly found in the anterior nares , throat , and skin it persistently colonises ~30% of the population with anywhere from 50–80% of individuals carrying it at any particular time point [3] . S . aureus is an opportunistic pathogen and although considered commensal , in many situations is capable of overcoming the host barrier defences to infect potentially any part of the body [4 , 5] . The ability of S . aureus to so effectively cause infection is a consequence of the myriad of virulence factors it produces . Toxins , enzymes , adhesion molecules , and immune evasion molecules allow the bacterium to first invade the host and then to get established and avoid destruction by the immune system [6 , 7] . One such class of virulence factor , the superantigen ( SAg ) , has been extensively studied over the past few decades ( reviewed in [8–18] . These toxins are known to cause staphylococcal toxic shock syndrome and staphylococcal food poisoning , and have been implicated in a number of conditions including sepsis , endocarditis , and pneumonia . The staphylococcal SAgs form a large family of related toxins with the SAgs found in many streptococcal species , most notably Streptococcus pyogenes . A critical SAg involvement in establishing infection of its natural niche has been discovered for S . pyogenes [19] , and a role for SAgs in promoting survival of S . aureus during infection has been identified [20] . The bacterial SAgs of S . aureus and S . pyogenes are a family of secreted toxins of around 20–30 kD that are most notably known for causing the toxic shock syndromes associated with these pathogens . By simultaneously binding major histocompatibility complex ( MHC ) class II on antigen presenting cells and the T cell receptor ( TcR ) in a Vβ-specific manner they are able to activate large proportions of T cells to produce pro-inflammatory cytokines . It is the ensuing ‘cytokine storm’ that is responsible for the symptoms of shock and organ failure that result . Currently there are 25 known staphylococcal and 11 streptococcal SAgs many of which have been crystallized alone or in complex with MHC class II , TcR , or both [21 , 22] . These SAgs share a common fold consisting of two highly stable domains , the N-terminal OB-fold domain and the C-terminal β-grasp domain , that are separated by a long , partially solvent-accessible central α-helix [23] . Interestingly , this same protein fold is also found in the functionally unrelated , 14-member family of Staphylococcal Superantigen-Like toxins ( SSLs ) [24] . Although the staphylococcal and streptococcal SAgs and the SSLs share limited sequence homology they can be identified by the two highly conserved PROSITE “family signature motifs” Y-G-G-[LIV]-T-X ( 4 ) -N ( Prosite entry PS00277 ) and K-X ( 2 ) -[LIVF]-X ( 4 ) -[LIVF]-D-X ( 3 ) -R-X ( 2 ) -L-X ( 5 ) -[LIV]-Y ( PS00278 ) [25] . The Staphylococcal Superantigen-Like ( SSL ) family of proteins are related to the SAgs by sequence and structure [26] . The SSLs however do not function as superantigens , rather are involved in blocking various aspects of host immunity . For instance , SSL7 binds to IgA and complement C5 to prevent C5a-mediated chemotaxis of inflammatory myeloid cells and C5-dependent microbial killing [27 , 28] . SSL10 binds to human IgG1 to inhibit the phagocytosis and complement activation mediated by this important immunoglobulin [29 , 30] . SSL3 binds to toll-like receptor 2 and inhibits its capacity to signal in response to pathogen associated molecular patterns [31 , 32] . SSL5 and SSL11 bind to P-selectin glycoprotein ligand-1 ( PSGL-1 ) via a highly conserved site with specificity for sialylated glycans that contain the minimal conserved trisaccharide , sialyl-lactosamine ( sLacNac = NeuAcα2-3Galβ1-4GlcNAc ) . The interaction with PSGL-1 prevents P-selectin mediated immune cell recruitment . [33 , 34] . They share this binding site with SSLs 2–4 , and SSL6 [35] . This conserved site has been implicated in the interactions of the sialylated-glycan binding SSLs with a diverse range of additional host glycoproteins that include the Fc receptor for IgA , the glycosylated N-terminal region of G protein-coupled receptors , platelet glycoproteins ( GPIbα , GPIIb-IIIa , and GPVI ) , CD47 , and matrix metalloproteinase-9 [34–40] . The recently discovered staphylococcal SAg , SElX , is unusual in that it is chromosomally located and thus found in all strains of S . aureus with the exception of clonal complex 30 [41] . It possesses an entirely unique N-terminus of no known homology that is much shorter in residues than the OB-fold domain of other SAgs . The C-terminal half of SElX however displays amino acid similarity with the β-grasp domain of both the SAgs and the SSLs . Recently SElX , like SSL5 and SSL11 , was shown to bind PSGL-1 in a sialylated-glycan-dependent manner to inhibit its interaction with P-selectin [38] . Here we present functional and structural evidence that SElX is an ‘SSL-like’ SAg . It has specificity for sLacNac and interacts with myeloid cells in a sialylated glycan-dependent manner to inhibit host defences . Additionally , X-ray crystallography of SElX reveals a unique structural variation from the typical SAg architecture , with the complete omission of an OB-fold domain .
SElX has been reported to display highest homology to TSST-1 and SSL7 [41] . Phylogenetic analysis indicates that SElX is more closely related to the SSL family of immune evasion proteins than to the bacterial superantigen family ( Fig 1A ) . Additionally SElX shows closer sequence conservation to the SSLs in a region of the central α-helix that makes up the PROSITE signature sequence PS00278 . In SElX the sequence KELD has higher identity to the SSL consensus sequence of KE ( L/I ) D than the consensus sequence of the SAgs QE ( L/I/V ) D . The Lysine ( K ) of this motif is absolutely conserved in the SSLs whereas the Glutamine ( Q ) is almost entirely conserved in the SAgs ( S1 Fig ) . Amino acid sequence alignment with the SSLs reveals that SElX possesses significant identity to SSLs 2–6 and SSL11 in the region of conservation that describes the glycan binding site of this SSL subfamily ( Fig 1B ) [34 , 35 , 42] . Of the seventeen residues that define the conserved glycan-binding site , SElX displays higher homology than SSLs from outside the glycan-binding subfamily . In particular , residues known to interact with sialylated glycans are highly conserved in SElX ( Fig 1B ) . Furthermore , no homology with residues that form the α-chain or β-chain MHC class II binding sites can be seen upon alignment of SElX with the other bacterial SAgs ( S1 Fig ) . For this reason the characterization of SElX as a novel SSL-like SAg was performed . Two variants of SElX were analysed , one being SElX2 cloned from the CC8 strain Newman and the other SElX8 cloned from strain JSNZ . JSNZ is a mouse-adapted strain of S . aureus isolated from preputial gland abscesses during a severe outbreak among male C57BL/6 mice [43] . This strain is from the multilocus sequence type ST88 and encodes no other identifiable SAgs other than SElX . To determine if SElX is a glycan binding protein , host protein binding assays were performed to compare SElX with the known glycan binding proteins SSL4 , SSL6 , and SSL11 . The glycan binding SSLs display broad binding capacity for plasma and myeloid cell glycoproteins [33–40 , 42] . Recombinant SElX2 , SSL6 , and SSL11 coupled to sepharose were used to isolate interacting proteins from human peripheral blood mononuclear cells ( PBMC ) , peripheral blood polymorphonuclear cells ( PMN ) , platelets , and plasma . Binding to mouse serum , and splenocyte and bone marrow ( BM ) derived cell lysates was also compared . The binding profiles of proteins bound by SElX2 in each instance were very similar to those of SSL6 and SSL11 ( Fig 2A ) . Previously , mutagenesis of the conserved Threonine ( T ) or Arginine ( R ) ( indicated by the red asterisks in Fig 1B ) in the glycan binding SSLs has resulted in greatly diminished capacity for carbohydrate-dependent interactions [34 , 35 , 42] . Mutation of the equivalent residue Threonine 130 ( T130 ) or Arginine 141 ( R141 ) to Alanine in SElX2 resulted in a significant reduction in host protein binding that is similar to mutating the equivalent Threonine in SSL11 or Arginine in SSL6 ( Fig 2A and S2 Fig ) . Additionally , SElX displayed a comparable energy- and glycan binding site-dependent binding to neutrophils as has previously been reported for SSL4 and SSL11 [34 , 42] ( S2 Fig ) . Identification of host leukocyte proteins bound by SElX and its glycan-binding site mutant SElX-T130A/R141A was performed using liquid chromatography—tandem mass spectrometry ( LC-MS/MS ) on lysate proteins captured by affinity precipitation using the immobilized SElX variants . The top scoring proteins identified are shown in Fig 2B and full data on the identified proteins and their functions is available in S1 Table . The mass spectrometry data show that many of the leukocyte proteins that are bound by SElX are integrins . Other adhesion molecules such as P-selectin and PECAM-1 feature in this list . Several of the SElX-targeted proteins are cytoplasmic in origin and are predominantly granule proteins or are associated with the cytoskeletal network , with roles linking the cytoskeleton to surface receptors . Furthermore , a large number of the identified proteins are involved in coagulation . The predominance of proteins identified to interact with SElX-T1301/R141 were intracellular and associated with the cytoskeleton with the integrin alpha IIb and beta 3 chains the only exception . Flow cytometry of recombinant SElX2 conjugated to Alexa Fluor 448 ( SElX-448 ) revealed that SElX bound to human granulocytes , monocytes , and weakly to lymphocytes whereas negligible binding was seen using the glycan-binding site mutant SElX-T130A/R141A ( Fig 3A ) . To further support the glycan-binding site dependency of cell binding , competition for the binding of SElX-488 to neutrophils using increasing concentrations of SElX or SElX-T130A/R141A was performed . The cell surface interaction of SElX-488 could be inhibited in a dose dependent manner with SElX whereas the glycan-binding site mutant showed no significant inhibition at any of the concentrations used ( Fig 3B ) . The host specificity of SElX binding was compared using human peripheral blood leukocytes and mouse splenic and bone marrow leukocytes . The SElX8-488 variant from the mouse-adapted strain JSNZ was used for this analysis and showed a greater capacity for binding to human cells ( Fig 3C ) . Recombinant SElX2 was analysed for carbohydrate binding to a glycan array that contained 611 mammalian glycan targets by the Consortium for Functional Glycomics . The array screening confirmed that SElX2 bound glycans containing the trisaccharide sialyl-lactosamine ( sLacNac = Neu5Aca2-3Galb1-4GlcNAc ) . Several of the strongly bound glycans terminated in the sLacNac-containing tetrasaccharide sialyl Lewis X ( sLeX = Neu5Acα2-3Galβ1-4 ( Fucα1–3 ) GlcNAc ) ( Table 1 and S3 Fig ) . Surface plasmon resonance ( SPR ) was employed to study the interaction between SElX and its sialylated target ligands . The affinities of recombinant SElX2 and SElX8 for sLeX and its core trisaccharide sLacNac ( a subcomponent of sLeX lacking the terminal fucose ) were determined by passing a concentration series of the highly purified proteins over the immobilized carbohydrates ( Fig 4A ) . Equilibrium dissociation constants ( KD ) were calculated from the equilibrium binding curves acquired from both SElX2 and SElX8 binding to sLeX and sLacNac . The KD of SElX2 was determined to be 22 . 90 ± 0 . 13 μM for sLeX , and 23 . 17 ± 1 . 05 μM for sLacNac . SElX8 bound sLeX and sLacNac with KD’s of 9 . 58 ± 1 . 02 μM and 14 . 21 ± 2 . 57 μM , respectively . Negligible binding was observed for the T130A , R141A , and T130A/R141A mutants of SEIX2 and SEIX8 to either sLeX or sLacNac ( Fig 4B ) . The crystal structure of SEIX8 was determined in complex with sLeX . The protein structure was solved by molecular replacement with a partial model of SSL4 ( PDB: 4DXG ) using residues 130–200 and was refined at 1 . 66 Å ( Table 2 ) . The C-terminal domain of SEIX8 ( residues N61–I161 ) adopts the β-grasp fold typical of the classical SAgs and members of the SSL family . The first 21 residues and the last 3 residues of the mature protein were undefined in the crystal structure , lacking electron density . The first structured residue N22 defines the start of the α-helix that sits atop the rear of the β-grasp C-terminal domain ( Fig 5A ) . An extended loop , which is stabilized by extensive hydrogen bonds , links the bottom of this helix directly with the first β strand of the β-grasp domain . ( Fig 5A ) . This linker region is predominantly unstructured , containing a β-hairpin and a series of β-turns as it packs across the side of the β-grasp domain . It completely replaces the typical OB-fold domain , revealing SElX to be a unique single-domain SAg . The binding site for sLeX is a V-shaped depression in the side of the β-grasp domain , which is formed by residues from a β-strand , the opposing helix ( a 310-helix ) , and the irregular polypeptide loop that links them ( Fig 5A ) . This region is on the opposite side of the β-grasp to the N-terminus linker loop . Seven residues lining the sides of the depression hydrogen bond directly to sLeX ( Fig 5B ) . These are the side-chains of T130 , E132 , K135 , Q138 , N140 , and R141 , and the main chain carbonyl of K128 . K128 , T130 , and R141 form an extensive network of hydrogen bonds with the sialic acid ( S ) of sLeX while Y129 participates in a hydrophobic interaction with this moiety . E132 hydrogen bonds with the galactose ( G ) and fucose ( F ) of sLeX , and with the side chain of K135 , while K135 makes one additional hydrogen bond to the fucose . Extensive hydrogen bonding occurs between Q138 and both the galactose and N-acetylglucosamine ( GlcNAc or N ) sugars with further hydrogen bonding to GlcNAc provided by N140 . An additional three waters make intermediate contacts directly between sLeX and SEIX . R141 makes an extensive network of hydrogen bonds to T130 , Q138 , and V144 across the floor of the binding site . The most striking feature of SElX is the absence of the ubiquitous N-terminal SAg/SSL OB-fold domain . Structural comparison with TSST-1 and SSL5 reveals that despite this deleted domain , the N-terminal α-helix of SElX is spatially conserved ( Fig 5C ) and just like the SAgs and SSLs it packs against the back of the β-grasp domain . The sLeX binding site of SElX shows high structural conservation with the 17-residue glycan-binding region of the SSLs . The RMSDs ( all atoms ) are 0 . 73927 Å ( over 152 atoms ) , 0 . 900455 Å ( over 155 atoms ) , and 0 . 78852 Å ( over 154 atoms ) between SElX and SSL4 , SSL5 , and SSL11 , respectively ( Fig 5D ) . Indeed , the binding of sLeX is almost entirely conserved with the sLeX binding of SSL4 [42] , SSL5 [35] , and SSL11 [34] . The only exception is the loss of a conserved aspartic acid ( V144 in SEIX ) , which in the SSL’s interacts with the sialic acid moiety of the glycan ( Fig 5D ) . TSST-1 interacts with the TcRVβ chain via an interface that includes the back of the N-terminal α-helix , the central α-helix of the β-grasp domain , and the top of the OB-fold [44] . A structural overlay of SElX with the structure of TSST-1 in complex with TcRVβ2 reveals that this unique SAg has the potential to bind TcRVβ in a similar fashion to TSST-1 . The β-grasps of both proteins overlay very well as do their N-terminal α-helices ( Fig 6 ) . This structural overlay places the TcR β-chain in close proximity to SElX . In particular , the N-terminal α-helix , the end of the central α-helix and the linker loop region that connects the N-terminus of SElX with the β-grasp domain ( Fig 6 ) . The overlay described here shows that SElX has the potential to bind TcRVβ predominantly via its spatially conserved N-terminal and central α-helices and potentially compensates for the lack of any OB-fold-TcRVβ interaction by the proximity of its connecting loop region to the TcRVβ chain . It is also evident from this structural overlay that the positioning of the potential TcRVβ binding site is on the opposite face of SElX to its sialylated glycan binding site and also leaves the concave face of β-grasp exposed for additional host interactions . SAgs have two known binding sites for MHC class II . A hydrophobic ridge in the OB-fold domain allows for binding to the invariant MHC class II α-chain , while three conserved residues in the β-grasp domain participate in the tetravalent co-ordination of a zinc molecule with an invariant histidine on the polymorphic MHC class II β-chain [45–49] . Having no OB-fold domain means that SElX cannot interact with MHC class II via the traditional α-chain-binding site . Additionally , no structural homology to the zinc-co-ordinating residues is identifiable in the β-grasp of SElX . The ability of SElX to bind MHC class II was tested . SElX8 conjugated to sepharose was used to isolate MHC class II from cell lysates of the MHC class II ( HLA-DR1 allele ) -expressing cell line LG-2 [50] . Immunoblot analysis using a polyclonal antibody against HLA-DR1 showed SElX isolated bands consistent with those of the MHC class II alpha and beta chains recognised by the anti-DR1 antibody ( Fig 7A ) . MHC class II was also affinity isolated by TSST-1 conjugated to sepharose but negligibly isolated by SElX8-T130A/R141A . This confirms that there is a sialylated-glycan-dependent binding of SElX to MHC class II , rather than the traditionally identified SAg binding sites . To support this , TSST-1 bound MHC class II from both neuraminidase-treated and untreated LG-2 cells , whereas SElX only isolated MHC class II from untreated cells . Glycan binding did not appear to influence the superantigen activity of SElX however , since both SElX and SElX-T130A/R141A displayed an equivalent capacity to stimulate the proliferation of PBMCs ( Fig 7B ) . Both SElX and its glycan-binding site mutant displayed ½ maximal PBMC stimulation at approximately 10ng/ml , whereas TSST-1 had a more typical superantigenic potential with its ½ maximal stimulation in the low pg/ml range [21] . The contribution of SElX to the survival of S . aureus was studied using a whole blood killing assay designed to represent an in vitro model of bacteraemia . Deletion of selX from strain JSNZ ( JSNZΔselX ) resulted in a significant reduction in its ability to survive in human blood ( Fig 8A ) . Addition of recombinant SElX to the assay increased JSNZΔselX survival . Complementation with the glycan-binding site mutant SElX-R141A did not increase the survival of JSNZΔselX in human blood indicating that the protective effect of SEIX was dependent on the sialylated glycan-binding site . To further validate this , S . aureus JSNZ strains were generated in which the deleted selX gene was replaced with either a glycan-binding site mutant gene selXR141 ( JSNZselXR141A ) , or repaired by the re-introduction of selX ( JSNZselX-REP ) . Survival of the selX repaired strain , JSNZselX-REP , was comparable to wild type with significantly higher cfu’s recovered from both of these strains compared to JSNZΔselX . The mutant of S . aureus carrying the glycan-binding site mutated selX gene , JSNZselXR141A , showed a significant reduction in survival compared to wild type and did not provide any significant protective advantage over the selX deletion mutant from which it was derived ( Fig 8B ) . All the modified strains displayed comparable in vitro growth curves and immunoblot analysis confirmed that JSNZselX-REP and JSNZselXR141A both produced SElX ( S4 Fig ) . A comparison between S . aureus JSNZ and JSNZΔselX was made using mouse models of subcutaneous infection and systemic infection . No significant difference in survival was observed between the wild type bacteria and the SElX-deficient strain in either model ( S5 Fig ) . Analysis of the 5’ untranslated region of selx was conducted to look for regulators of selx expression . A recent study into the control of ssl1 , ssl7 , ssl9 , and ssl11 expression revealed promoter elements that include a binding site for SaeR , the response regulator of the S . aureus exoprotein expression ( Sae ) two component system ( TCS ) 5’ to the translational start site of these genes [51] . Comparison of the upstream regions of selx2 and selx8 with those of these ssls confirmed that the selx gene also possesses a direct repeat sequence with high identity to the conserved SaeR binding site [52 , 53] . This sequence , GTTAA ( n6 ) GTTAA is seen directly upstream from the -35 and -10 promoter elements . An additional ½ SaeR binding site was identified 6 bases further upstream . By performing 5’ RACE the transcription start site for selx was identified 7 bases 3’ to the entirely conserved Pribnow box ( or -10 promoter sequence ) ( Fig 9A ) . Using a mouse model of subcutaneous infection , the expression levels of selx transcripts present in S . aureus strain Newman and JSNZ abscesses at 24 , 48 , and 96 hours after infection were compared with that of in vitro culture of the inoculum . During S . aureus Newman subcutaneous infection , an up-regulation of expression was observed from the earliest time point that was approximately 30-fold greater than the in vitro culture at the time of infection ( Fig 9B ) . A corresponding trend towards increased saeR transcript expression was also observed at day one that declined in conjunction with that of selx expression . Expression of selx during subcutaneous infection of mice with S . aureus JSNZ was increased over 100-fold at 24 hrs in relation to the inoculum and remained elevated over the four day study . A robust and sustained increase in saeR was also measured over this time period ( Fig 9B ) . In support of selx regulation by Sae , immunoblot analysis of the culture supernatant from an Sae TCS-deficient strain of S . aureus revealed the complete absence of any SElX production ( S4 Fig ) .
The functional and structural insights gained from this research reveal that SElX is a unique member of the SAg family that shares features of the related SSL family . Though it was previously described as a SAg [41] , we have shown that it also possesses the conserved sialylated-glycan binding site of the SSLs . It shows similar host factor binding profiles to the SSLs and mutation of key conserved residues in the binding site greatly reduces this toxins binding capacity for host proteins . The specificity of SElX , as determined by glycan array screening , was comparable to that of the SSLs , with SPR analysis confirming its affinity for sLacNac and sLeX . The affinity of SElX for sLacNac ( 14 μM for SElX8 and 23 μM for SElX2 ) is lower than the SSLs which range from 0 . 47 μM for SSL4 to 2 . 4 μM for SSL11 [42] . Perhaps this is a consequence of it lacking a glycan-binding aspartic acid that is conserved in the SSLs . SElX possesses a valine at this position that does not interact with sLeX in the crystal structure . The co-crystallization of SElX with sLeX did show however that all the remaining residues that interact with the tetrasaccharide are conserved with those of the SSLs . Many of the leukocyte proteins bound by SElX were adhesion molecules including P-selectin , PECAM1 and notably several integrin’s that have important roles in immune recognition and cell activation . It is evident from the annotation of protein function ( S1 Table ) that several of the proteins interacting with SElX are involved in coagulation . By binding to these molecules , SElX has the capacity to interfere with important host immune and wound healing functions during infection to help the bacteria evade destruction . When comparing the proteins identified by affinity interaction with SElX-T130A/R141A to those bound by SElX , it is clear that the interaction with cell surface receptors has been lost , with the exception of integrin alpha IIb and beta 3 . The top scoring identified proteins bound by this glycan-binding site mutant are predominantly intracellular and cytoskeletally-related . It is not possible to determine from our approach which of these host proteins bind directly or indirectly to SElX so it is conceivable that some of these proteins are isolated in complex with glycoproteins that directly bind to SEIX . SElX may interact with cytoskeletal components or they may be present as a consequence of their connection with directly bound surface receptors . If SElX does bind cytoskeletal structures this would give it the potential to interfere with cellular rearrangements that affect cell movement , phagocytosis , receptor recycling , and even the release of granule contents . SElX may possess both glycan-dependent and -independent binding sites . The affinity precipitation assays showed that the glycan-site mutants retained some protein binding capacity . These targets may be intracellular proteins based on the mass spectrometry data and the observation that the glycan-binding site mutant of SElX displayed negligible cell surface binding by flow cytometry . The nature of these SElX-host interactions is a focus of further investigation . We used the WBK assay as an in vitro representation of staphylococcal bacteraemia and the relevance of using whole blood for identifying correlates of S . aureus virulence has been highlighted recently [54–56] . This assay revealed that SElX was necessary and sufficient for successful survival of S . aureus in human blood . Removal of selX caused a reduction in bacterial load of almost 1-log ( or 85% ) that could be fully restored by the addition of SElX or the reinstatement of selX . What’s more , the protection afforded by SElX was determined to be reliant on its glycan-binding site and reveals a unique property that is more in keeping with immune evasion by the glycan-binding SSLs than superantigenic immunomodulation . We found the consensus direct-repeat binding sequence for SaeR , the response regulator of the saeRS two-component regulatory system , in the immediate region upstream of selx . In support of this regulator being involved in the control of selx expression , we found there to be a complete lack of SElX produced by a Sae TCS-deficient mutant of S . aureus . This indicates that , like the SSLs , expression of SElX would be turned on in situations when the bacterium comes under stress such as from attack by host defence mechanisms [57–60] . SElX would likely be produced at the same early time point as the SSLs to perform complimentary functions in immune evasion . The protection provided by SElX to the survival of S . aureus in the whole blood killing assay supports its role in blocking host innate defences . The sustained expression of selx over the 96 hour course of murine subcutaneous infection seen in S aureus JSNZ may be a consequence of the more robust in vivo expression of saeR in this strain compared to that observed from Newman . It should be noted however that the Sae TCS is constitutively active in Newman [58 , 61] and that could explain the lower relative increase in selx up-regulation in vivo . Indeed we observed that Newman produces more SElX in vitro than JSNZ ( S4 Fig ) . Despite the ability to bind mouse cellular and serum proteins , and up-regulation during murine subcutaneous infection , we did not see a significant effect of SElX when comparing the selX gene deletion mutant with wild type S . aureus in the murine models of infection . One reason for this may be the weaker binding capacity of SElX for mouse cells . It is also possible that the sialylation of receptors in the mouse may be sufficiently different in pattern and/or magnitude to preclude SElX from interacting with the same set of glycoproteins that it does with the human host [62 , 63] . It has previously been reported that SElX does not contribute to disease severity in a murine model of pneumonia [64] . However , by deleting selX in S . aureus an attenuation of virulence has been observed in both a rabbit model of necrotizing pneumonia [41] and a bovine model of mastitis [65] suggesting that alternate species to mice are more appropriate for investigating the contribution of SElX during disease . X-ray crystallography revealed the unique single-domain structure of SElX . It completely lacks an OB-fold domain and its spatially conserved N-terminal α-helix is linked to the β-grasp domain by a short unstructured region . In addition to the SSLs , S . aureus produces other β-grasp domain exoproteins that are involved in immune evasion . These include the Chemotaxis Inhibitory Protein of S . aureus ( CHIPS ) and the Formyl Peptide Receptor-like 1 Inhibitory Protein ( FLIPr ) [25 , 66] . Both of these immune evasion molecules display conservation of sequence and structure in the region of the glycan binding site [25] . Being a single-domain SAg means that it would not be possible for SElX to interact with TcRs in the same manner as most of the traditional SAgs that typically bind the TcRVβ chain via an interface involving the cleft between the N-terminal α‐helix and the top of the OB‐fold [67 , 68] . Rather it has the potential to bind the TcR in a similar fashion to TSST-1 by using predominantly its two α-helices and its novel linker region , suggesting that SElX has maintained the minimal requirement at its N-terminus for engaging the TcR . However , because the first 21 residues are not defined in the structure , they cannot be precluded from providing additional contacts with the TcRVβ . Further investigation by co-crystallization with TcRVβ and mutagenesis will confirm this . SElX exhibits no conservation with residues of the SAgs involved in MHC class II binding . We found that the interaction of SElX with MHC class II was significantly influenced by its glycan-binding site . Yet this did not correlate with its ability to stimulate the proliferation of PBMCs since SElX-T130A/R141A displayed an equivalent activity to SElX . The superantigenic capacity of SElX however , was less potent than that of the typical SAgs [21] . These observations suggest a different mode of action for the T cell stimulation activity observed for SELX . This is currently under investigation . SElX is present in most S . aureus and is considered to have been acquired by an ancestor of the S . aureus species [41] . We have shown here that SElX possesses functions of two major related families of staphylococcal virulence factors , although it is less potent than typical SAgs and has weaker affinity for sialylated glycans compared to the SSLs . Perhaps SElX represents the missing-link between the related SAgs and SSLs and is the descendent of an ancestral staphylococcal virulence factor that had the properties of both these present-day families . While subsequent duplications of this ancestor led to the functional specialization seen in the SAgs and SSLs , SElX instead evolved to retain the functional properties of the precursor protein , losing its OB-fold domain in the process . Consequently , SElX targets both the adaptive immune system as a SAg and innate immune defences as an SSL . It is therefore unsurprising that this ‘SSL-like SAg' has been almost universally retained by all S . aureus . With the increasing seriousness of antimicrobial resistance associated with S . aureus , there is a need for alternate therapeutic interventions . The discovery of virulence factors like SElX that correlate with disease and the determination of their modes of action will allow for a more targeted approach to the development of anti-infectives that can be used to treat or prevent staphylococcal disease .
Blood was collected from healthy human volunteers who had given informed consent in writing in accordance with the University of Auckland Human Participants Ethics Committee ( UAHPEC ) guidelines . Animals were housed and cared for in accordance with The Animal Welfare Act ( 1999 ) and institutional guidelines provided by the University of Auckland Animal Ethics Committee , which reviewed and approved these experiments under application R847 . The mice were shaved and inoculated with S . aureus under isoflurane anaesthesia . They were monitored daily for alterations in body weight and general health . Mice were euthanized by CO2 inhalation . The gene for TSST-1 was cloned from S . aureus strain RC31187 and the various SElX and SSL6 genes and mutants from S . aureus strain Newman or S . aureus strain JSNZ using the primers listed in Table 3 . Mutants were generated by overlap PCR using internal overlapping primers containing the mutation together with the external cloning primers . The genes were cloned into pET32a-3C , recombinant proteins were expressed in Escherichia coli AD494 ( DE3 ) pLysS as thioredoxin fusion proteins , and isolated by nickel affinity chromatography ( Ni Sepharose 6 Fast Flow , GE Healthcare ) . The thioredoxin was cleaved off with 3C protease and the proteins were further purified using ion-exchange chromatography ( SElX , SSL6—MonoS , SSL11 , TSST-1—MonoQ . GE Healthcare ) . Production of SSL11 has been described previously [34 , 42] . Recombinant proteins were coupled to sepharose in accordance with the manufacturer’s guidelines . Protein at 2 mg/ml in PBS pH 8 . 0 was added to cyanogen bromide activated sepharose ( GE Healthcare ) to a final concentration of approximately 5mg protein/ml sepharose . The slurry was incubated with slow inversion at room temperature until a negligible amount of protein remained in the supernatant . Any remaining active sites on the sepharose were quenched by incubation in 100mM Tris pH8 . 0/150mM NaCl for 2 hours at room temperature before the sepharose was repeatedly washed with PBS pH 8 . 0/0 . 1% sodium azide and stored at 4°C as a 1:1 slurry in PBS/azide . Fresh human blood was collected in Heparin vacutainer tubes ( BD Biosciences ) . Granulocytes , mononuclear cells and plasma were isolated by separation through a Histopaque 1077 over Histopaque 1119 ( Sigma ) double density centrifugation gradient according to the manufacturer’s instructions . LG-2 cells were prepared by culturing in complete RPMI-1640/10%FCS ( Gibco ) . For neuraminidase treatment , 1x107 LG-2 cells/mL in 150 mM NaCl , 5 mM CaCl2 pH 6 . 0 were incubated with or without 25 unit/ml neuraminidase ( New England Biolabs ) for 1 h at 37° in a 5% CO2 incubator with occasional mixing . To isolate mouse leukocytes , female BALB/c mice aged 5–6 weeks were culled via CO2 asphyxiation and the spleens , tibiae and fibulae were removed . Single cell suspensions of splenocytes were created by pushing the tissue through a sterile metal sieve . Any remaining cells still present on the sieve were washed through with sterile PBS . Bone marrow was flushed from cut bones with PBS using a 20-gauge needle to create single cell suspensions which were filtered through a 70 μm strainer . Red blood cells were removed by passing the cell suspensions through a Histopaque 1083 ( Sigma ) gradient . The cells were then washed with PBS . Cell lysates were prepared by incubating 1x107 cells/ml in 10 mM Tris-HCl ( pH 8 . 0 ) , 140 mM NaCl , 1% ( v/v ) Triton X-100 , 1 mM iodoacetic acid , 1 mM PMSF , 0 . 025% NaN3 for 1 hr at 4°C , before centrifugation at 20000g for 30 min . 10 μl of protein:sepharose slurry was added to 100 μl of cell lysate or plasma in a total volume of 0 . 5 ml 10 mM Tris-HCl ( pH 8 . 0 ) , 140 mM NaCl , 1 mM PMSF , 0 . 025% NaN3 and incubated with slow inversion for 30 min at room temperature . The protein:sepharose was washed three times in 10 mM Tris-HCl ( pH 8 . 0 ) , 140 mM NaCl , 1% TX-100 , 1 mM PMSF , 0 . 025% NaN3 before being boiled in 2X sample buffer . The sample was separated by SDS-PAGE and visualized by Coomassie Blue staining or transferred to nitrocellulose membranes for Immunoblot analysis . Membranes were blocked at 4°C O/N in 10 mM Tris ( pH 8 . 0 ) , 120 mM NaCl , 0 . 1% Tween-20 , 5% Non-fat milk powder , probed for 1 hour at room temperature with rabbit anti-SElX polyclonal IgG ( made in-house ) or rabbit anti-huDR1 polyclonal IgG ( made in-house ) , followed by 1 hr with goat anti-rabbit IgG-HRP ( Dako ) . Chemiluminescence was performed using SuperSignal West Pico Chemiluminescent Substrate ( Thermo Scientific ) and images were captured using a LAS-3000 imager with Image Reader software ( Fujifilm ) . Heparinised blood was incubated in erythrocyte lysis buffer ( 150 mM NH4Cl , 10 mM KHCO3 , 0 . 1 mM EDTA , pH 7 . 4 ) and washed 3 times in PBS . The isolated leukocyte fraction was lysed at 1x107 cells/ml in 10mM Tris-HCl ( pH 8 . 0 ) , 140mM NaCl , 1% ( v/v ) Triton X-100 , 1mM iodoacetic acid , 1mM PMSF , 0 . 025% NaN3 for 1hr at 4°C prior to clarification by centrifugation at 20000xg for 30 min . 10μl of protein:sepharose slurry was added to 100μl of cell lysate in a total volume of 0 . 5ml 10mM Tris-HCl ( pH 8 . 0 ) , 140mM NaCl , 1mM PMSF , 0 . 025% NaN3 and incubated with slow inversion for 30 min at room temperature . The protein:sepharose was washed three times in 10mM Tris-HCl ( pH 8 . 0 ) , 140mM NaCl , 1% TX-100 , 1mM PMSF , 0 . 025% NaN3 . The protein:sepharose samples were incubated in 50μl elution buffer ( 6M urea/2M thiourea , 20mM tris ( pH 8 . 0 ) , 20mM NaCl , 5mM DTT ) for 30 min at room temperature with frequent mixing . The samples were centrifuged and the supernatant removed with a Hamilton syringe . A further 20 μl of elution buffer was added to the beads and incubated for 10 min . The supernatants from the two elutions were combined and centrifuged again . 50 μl was taken out , snap frozen in ethanol-dry ice and stored at -80C for mass spectroscopy analysis . 10 μl of the remainder was mixed with 10 μl 2x sample buffer and 10μl of this was analysed by SDS-PAGE . Protein identification was performed by LC-MS/MS using a Sciex TripleTOF 6600 by the Mass Spectrometry Centre , Auckland Science Analytical Services , The University of Auckland , Auckland , New Zealand . The data was analysed using ProteinPilot 5 . 0 . Recombinant SElX or SElX-T130A/R141A were coupled to Alexa Fluor 488 based on the manufacturer’s recommendations . One twentieth volume of 10 mg/ml Alexa Fluor 488 ( Life Technologies ) in DMSO was added to 10 mg/ml SElX in 0 . 1 M NaHCO3 pH 8 . 3 and incubated in the dark at room temperature for 2 hours . SElX conjugated with Alexa Fluor dye ( SElX-488 and SElX-T130A/R141A-488 ) were separated from free label using a HiTrap Desalting column ( GE Healthcare ) in PBS pH7 . 4 . Heparinised whole human blood was incubated in erythrocyte lysis buffer . The remaining leukocytes were washed in PBS and suspended at 1 x107 cells/ml in FACs buffer ( PBS/1% BSA ) . A two-fold dilution series of SElX-488 from 500nM was incubated with 1 x106 cells at room temperature for 15 min . For competition assays , 100 nM SElX-488 was added to 1 x106 cells with or without addition of 100 , 500 , or 1000 nM unlabelled SElX or SElX-T130A/R141A and incubated at room temperature for 15 min . Mouse cells were processed as described above . 1x106 mouse cells +/- 100nM SElX-488 were incubated in FACs buffer for 15 min with appropriate antibodies to identify the following populations: CD3-PE-Cy5 to identify T cells in the spleen; B220-Cytochrome to identify B cells in the spleen and bone marrow; and Gr-1-APC-Cy7 to identify myeloid cells , predominantly neutrophils , in the bone marrow . Cells were washed in FACs buffer prior to acquisition using a BD LSR II Flow Cytometer with FACsDiva ( BD Biosciences ) . Data analysis was performed using FlowJo ( FlowJo , LLC ) with cell populations gated as granulocytes , monocytes , and lymphocytes based on size and granularity , or by cell-specific marker expression . Analysis of human leukocyte binding was performed in triplicate using 3 healthy individuals . The mouse cell binding was performed twice using n = 1 mouse each repeat . Statistical analysis was performed using Graphpad Prism . Live cell imaging was performed as previously described [42] . Briefly , 1×105 neutrophils were adhered to L-lysine coated glass bottom dishes ( World Precision Instruments ) for 30 min at RT in PBS pH 7 . 4 . 0 . 2 μM SElX-488 was incubated with the cells for 15 min at either 4°C or 37°C in PBS pH 7 . 4 . Excess SElX-488 was washed away with PBS pH 7 . 4 before being viewed by the Olympus FV1000 confocal scanning microscope at 600× magnification . The analysis software used was Olympus Fluoview v1 . 7b with resizing performed by ImageJ v1 . 46 . SElX-488 was sent to the Consortium for Functional Glycomics ( CFG ) : Protein-Glycan Interaction Core ( http://www . functionalglycomics . org/static/consortium/resources/resourcecoreh . shtml ) for screening of their mammalian glycan array . Binding was analysed at 100 , 200 , and 500 μg/ml to version 5 . 0 of the printed array consisting of 611 glycans in replicates of 6 . Relative binding was measured as relative fluorescent units ( RFU ) . The average RFU value from the replicates , the standard deviation , and %CV ( %CV = 100 X Std . Dev / Mean ) were calculated after removing the highest and lowest values from each set of 6 . The SElX-488 used for glycan screening is referred to as SSL0 on the consortium website and data from the screening can be found via the following link: ( http://www . functionalglycomics . org/glycomics/search/jsp/result . jsp ? query=ssl0&cat=all ) . Biosensor analysis of SElX interactions with sLeX and sLacNac were performed on a Biacore T200 ( GE Healthcare , Uppsala , Sweden ) . Ligands were coupled using carbodiimide chemistry to a CM5 biosensor chip surface according to manufacturer’s instructions . BSA-sLeX and BSA-sLacNac ( Dextra Laboratories ) were coupled at 200–250 RU in 100 mM Na-Formate pH4 . 3 . Remaining sites were blocked with BSA . Control channels for subtraction of bulk and non-specific responses were coupled with BSA to similar levels as test channels . SElX2 and the trailing edge from size exclusion chromatography of SElX8 ( concentration series from 50–0 . 25 μM ) , in HBS-EP+ ( 0 . 01 M HEPES pH7 . 4 , 0 . 15 M NaCl , 3 mM EDTA , 0 . 05% Surfactant P20 , GE Healthcare , Uppsala , Sweden ) , were passed over the immobilised ligands at 30 μl/min . The response at equilibrium ( Req ) was measured as the binding response plateau at 5 min . Surfaces were regenerated between cycles with 4 M GuCl . Equilibrium binding data were fitted to a steady state single binding site model using the Biacore T200 Evaluation software ( GE Healthcare ) . For comparison with binding site mutants , SElX proteins ( 20 μM ) were passed over the immobilised ligand at 30 μl/min for 100 s . Sensorgram overlays were carried out using the Biacore T200 Evaluation software ( GE Healthcare ) . Each experiment was performed in duplicate and repeated three times . The affinity ( KD ) values are expressed as mean ±SD of the repeats . Proliferation assays were performed in 96 well U bottomed plates . Peripheral Blood Mononuclear Cells ( PBMCs ) isolated by Histopaque 1077 ( Sigma ) density centrifugation were suspended at 1x106 cells/ml in complete RPMI-1640/10% FCS ( Gibco ) and then added in an equal volume to a 10-fold dilution series ( in triplicate ) of toxin starting from 20μg/ml in complete RPMI-1640/10% FCS ( Gibco ) . The plates were incubated at 37°C in 5% CO2 for 3 days . 0 . 25 μCi 3H-thymidine was added to each well and the plates incubated for a further 18hr . The plates were harvested to filter mats and incorporation of 3H-thymidine into cellular DNA was determined using a Wallac Jet 1450 Microbeta Trilux liquid scintillation counter ( Wallac ) . To generate JSNZΔselx the flanking regions of selx were amplified by PCR from JSNZ genomic DNA using the primers selX-upper-for with selX-upper-rev , and selX-lower-for with selX-lower-rev ( Table 3 ) . The upper and lower flanking region PCR products were then mixed and used as template for PCR with the selX-upper-for and selX-lower-rev primers . The resulting product was cleaved at the primer-introduced restriction sites , ligated with pIMAY cleaved with the same endonucleases , and transformed into E . coli DC10B . After sequence confirmation the plasmid isolated from DC10B was electroporated into JSNZ . Integration of pIMAY into JSNZ and excision of selX was performed as described by Monk et . al . 2012 [69] . Confirmation of gene deletion was confirmed by sequencing with the selX-outer-for and selX-outer-rev primers ( Table 3 ) . To generate JSNZselxR141A , JSNZ genomic DNA was amplified using selX-upper-for primer with the selXj-R141A-rev primer , and selX-lower-rev primer with the selXj-R141A-for primer ( Table 3 ) . The two PCR products were mixed and used as template for amplification using the selX-upper-for and selX-lower-rev primers . To generate JSNZΔselx-REP , JSNZ genomic DNA was amplified using selX-upper-for primer with the selX-REP-rev primer , and selX-lower-rev primer with the selX-REP-for primer ( Table 3 ) . The overlapping REP primers were designed to introduce a single synonymous substitution into the selx gene [41] . The two PCR products were mixed and used as template for amplification using the selX-upper-for and selX-lower-rev primers ( Table 3 ) . The resulting products were introduced into pIMAY using the primer-introduced restriction sites and transformed into DC10B . Following sequence confirmation each plasmid was electroporated into JSNZΔselx , allelic exchange was performed as described [69] , and introduction of selXR141A or selX-REP was confirmed by sequencing . Overnight cultures of S . aureus JSNZ , JSNZΔselx , JSNZselxR141A , or JSNZselx-REP in tryptic soy broth were diluted 1/100 and cultured at 37°C until mid log-phase . After suspension in Hanks balanced salt solution to an OD600 of 0 . 4 ( = ~1x108 cells/ml ) , 1x105 CFU S . aureus , JSNZΔselx , JSNZselxR141A , or JSNZselx-REP were incubated with 70% whole blood with or without recombinant SElX at the indicated concentrations for 20 hr at 37°C with gentle shaking . Dilutions of the suspensions at time 0 and after 20 hr were plated in triplicate onto tryptic soy agar and incubated O/N at 37°C for enumeration . Each assay was performed in duplicate with enumerations made in triplicate on at least three individual donors . Statistics were performed using Graphpad Prism . Kruskal—Wallis one way analysis of variance ( ANOVA ) was performed and comparisons between samples were made using Dunn's Multiple Comparison Test . Subcutaneous infection of mice with S . aureus Newman or JSNZ was performed as previously described [43] . Log-phase S . aureus were washed and diluted in PBS and then mixed 1:1 in a sterile cytodex bead ( Sigma ) solution ( 0 . 5 g/ml in PBS ) . Female CD1 mice aged 7–8 weeks were anaesthetized with isoflurane , the flank area shaved and 5x106 CFU bacteria was injected subcutaneously into the flank . Mice were euthanized by CO2 inhalation and abscess tissue ( from 2 independent experiments containing n = 3 mice per treatment group ) was aseptically collected from groups of mice at 24 , 48 , and 96 hours post infection . For transcript analysis , 1 ml RNAprotect Bacteria reagent ( Qiagen ) was added to each abscess immediately upon excision . The samples were pelleted , suspended in a further 1 ml of RNAprotect Bacteria reagent and pelleted again . The pellets were suspended in 0 . 3 ml TE buffer containing 25 μg lysostaphin ( Sigma ) and incubated for 1 hr . Seven 0 . 1 mm silica/zirconia beads ( Omni ) and 1ml Trizol LS reagent ( LifeTech ) were added and the samples beaten in an Omni BeadRupter-24 . After addition of chloroform the samples were centrifuged and the extracted RNA was ethanol precipitated . Contaminating DNA was removed using Turbo DNase ( Ambion ) and the bacterial RNA was enriched for using a Microbenrich kit ( Ambion ) and amplified using a MessageAmp II Bacteria kit ( Ambion ) . RNA was extracted from the inoculum as for the abscess samples ( with exclusion of the Microbenrich step ) to provide for in vitro comparisons . Complementary DNA was synthesized from total RNA using Superscript lll first strand synthesis supermix ( LifeTech ) and stored at -80°C . Real time PCR analysis was performed in triplicate using an Applied Biosystems 7900HT Fast Real-Time PCR System with PerfeCTa SYBR Green FastMix ROX ( Quanta Biosciences ) on 1/2 dilutions of the synthesized cDNA and selX-specific primers ( Table 3 ) . Real time PCR data was normalised against the reference genes ftsZ and gyrβ . Arbitrary gene expression values were converted to ratios against an average of the in vitro data . The identification of the selX transcriptional start site was achieved using the method described by Miller at . al . 2015 [70] . Briefly , first strand cDNA was synthesised from S . aureus RNA using selX-GSP1 ( Table 3 ) and SuperScript RT III/RNAse First Strand Synthesis Mix ( Life Technologies ) according to the manufacturer’s instructions . After incubation at 70°C to terminate the reaction , the cDNA was incubated at 37°C with RNase H ( Thermo Fisher Scientific ) , then column purified ( Zymo Research ) . A poly-C tail was added to the 3’ end of the cDNA using Terminal Deoxynucleotidyl Transferase ( Thermo Fisher Scientific ) and then amplified using selX-GSP2 and the Abridged anchor primer ( Table 3 ) . A nested amplification round of PCR was performed on the product using selX-GSP3 and the Abridged universal amplification primer ( Table 3 ) . The nested product was isolated by agarose gel electrophoresis and extracted using a NucleoSpin Gel and PCR clean-up kit ( Macherey-Nagel ) . The product was cloned into pBluescript using the restriction enzymes EcoRI and SalI and the plasmid from six positive transformants were sequenced . SEIX ( 10 mg/ml ) was co-crystallized with 5 mM sLeX ( Dextra Laboratories ) in the presence of 24% polyethylene glycol ( PEG ) 3350 and 250 mM Tri-Lithium citrate pH 7 . 5 at 17°C . Protein crystals formed within 7 days . The protein crystal was flash-cooled in the same crystallization condition supplemented with 20% glycerol . Diffraction data was collected at 1 . 5418 wavelength after 2x 1 sec room temperature annealing . The protein structure of SEIX8 in complex with sLeX was solved by molecular replacement with a partial model of SSL4 ( PDB: 4DXG ) residues 130–200 and was refined at 1 . 66 Å using Phaser MR and REFMAC in CCP4 suites ( Table 2 ) . Structural comparisons of the glycan binding sites were made using LSQ in coot for all atom rmsd . The atomic coordinates and structure factors for SElX complexed with sLeX have been deposited in the PDB under code no . 5U75 . | The ability of Staphylococcus aureus to cause disease can be attributed to the wide range of toxins and immune evasion molecules it produces . The 25-member superantigen ( SAg ) family of toxins disrupts adaptive immunity by activating large proportions of T cells . In contrast , the structurally-related 14-member Staphylococcal Superantigen-Like ( SSL ) family inhibits a wide range of innate immune functions . We have discovered that the SAg staphylococcal enterotoxin-like X ( SElX ) has the sialylated-glycan-dependent active site found in a sub-family of SSLs . Through this site it possesses the ability to affect host innate immunity defences . By solving the X-ray crystal structure of SElX we have also discovered that SElX is a unique single-domain SAg . While it retains a typical β-grasp domain , it lacks the OB-fold domain that is present in all other staphylococcal SAgs . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
]
| [
"cell",
"binding",
"cell",
"physiology",
"medicine",
"and",
"health",
"sciences",
"chemical",
"characterization",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"staphylococcus",
"aureus",
"animal",
"models",
"clinical",
"medicine",
"model",
"organisms",
"experimental",
"organism",
"systems",
"sequence",
"motif",
"analysis",
"bacteria",
"bacterial",
"pathogens",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"major",
"histocompatibility",
"complex",
"bioinformatics",
"staphylococcus",
"medical",
"microbiology",
"microbial",
"pathogens",
"mouse",
"models",
"binding",
"analysis",
"blood",
"cell",
"biology",
"clinical",
"immunology",
"anatomy",
"physiology",
"database",
"and",
"informatics",
"methods",
"biology",
"and",
"life",
"sciences",
"organisms"
]
| 2017 | Staphylococcal enterotoxin-like X (SElX) is a unique superantigen with functional features of two major families of staphylococcal virulence factors |
Many genes of large double-stranded DNA viruses have a cellular origin , suggesting that host-to-virus horizontal transfer ( HT ) of DNA is recurrent . Yet , the frequency of these transfers has never been assessed in viral populations . Here we used ultra-deep DNA sequencing of 21 baculovirus populations extracted from two moth species to show that a large diversity of moth DNA sequences ( n = 86 ) can integrate into viral genomes during the course of a viral infection . The majority of the 86 different moth DNA sequences are transposable elements ( TEs , n = 69 ) belonging to 10 superfamilies of DNA transposons and three superfamilies of retrotransposons . The remaining 17 sequences are moth sequences of unknown nature . In addition to bona fide DNA transposition , we uncover microhomology-mediated recombination as a mechanism explaining integration of moth sequences into viral genomes . Many sequences integrated multiple times at multiple positions along the viral genome . We detected a total of 27 , 504 insertions of moth sequences in the 21 viral populations and we calculate that on average , 4 . 8% of viruses harbor at least one moth sequence in these populations . Despite this substantial proportion , no insertion of moth DNA was maintained in any viral population after 10 successive infection cycles . Hence , there is a constant turnover of host DNA inserted into viral genomes each time the virus infects a moth . Finally , we found that at least 21 of the moth TEs integrated into viral genomes underwent repeated horizontal transfers between various insect species , including some lepidopterans susceptible to baculoviruses . Our results identify host DNA influx as a potent source of genetic diversity in viral populations . They also support a role for baculoviruses as vectors of DNA HT between insects , and call for an evaluation of possible gene or TE spread when using viruses as biopesticides or gene delivery vectors .
The genomes of large eukaryotic double-stranded DNA viruses contain high proportions of cellular genes resulting from host-to-virus horizontal transfers ( HT ) [1–4] . For example , at least 10% of giant virus genes and up to 30% of herpesvirus genes likely originated from eukaryote or prokaryote genomes [1 , 5 , 6] . Some of these genes have been shown to act as key factors in the etiology of viral diseases [7 , 8] . Because cellular gene content can be quite different between closely related viruses and/or quite similar between distantly related viruses [1 , 2 , 9] , viral co-option of host genes appears to be rather frequent during virus evolution . The cellular genes that have so far been identified in viral genomes result from relatively ancient host-to-virus HT events . From a population genetics perspective , these viral-borne host genes must have been inherited at low to intermediate frequencies over multiple rounds of viral replication until they finally reached fixation in the viral species , likely because they provided a fitness gain to the virus . In agreement with this hypothesis , many of these genes are thought to play a role in thwarting host anti-viral defenses , thus facilitating viral replication [10] . A corollary of this scenario is that many viral-borne host genes resulting from host-to-virus HT should be found at varying frequencies in viral populations . However , host-to-virus HT has never been investigated at the micro-evolutionary scale of the viral population . Therefore , the frequency of host-to-virus HT as well as the evolutionary and molecular processes underlying the capture and domestication of eukaryotic genes by viruses remain poorly understood . Another outstanding question arising from host-to-virus HT is whether viral-borne host genes acquired from a given host individual can be transferred to the genome of another infected individual through virus-to-host HT . In other words , can viruses act as vectors of HT between their hosts ? Hundreds of HT cases have been characterized in eukaryotes [11 , 12] . Many of these transfers have generated evolutionary novelties and allowed receiving organisms to adapt to new environments [13–15] . Horizontal transfer of DNA is therefore increasingly appreciated as an important evolutionary force shaping eukaryote genomes . However , the mechanisms and the potential vectors involved in HT of DNA between eukaryotes remain poorly known , especially in multicellular eukaryotes . Viruses have been proposed as candidate vectors facilitating HT between eukaryotes because they are transmitted horizontally ( and in some cases vertically ) and they replicate inside host cells [16 , 17] . In metazoans , the vast majority of HTs characterized so far are transfers of transposable elements ( TEs ) , which constitute pieces of DNA that are capable of moving from one genomic locus to another , often duplicating themselves in the process [18] . Several studies have uncovered host TEs packaged in viral capsids or even integrated into viral genomes , suggesting that TEs can jump from host to virus during the course of a viral infection [19–24] . We discovered two such TEs from the cabbage looper moth ( Trichoplusia ni ) integrated at low frequency in genome populations of the baculovirus Autographa californica multiple nucleopolyhedrovirus ( AcMNPV ) following infection of T . ni caterpillars [23] . Importantly , these two TEs show signs of HT between several sympatric moth species that can be infected by baculoviruses in the wild . The search for T . ni sequences integrated into populations of AcMNPV was however restricted to the dozen of T . ni genes and transposable elements that were known at the time . Therefore , the number and diversity of moth TEs and non-TEs that become integrated into AcMNPV genomes during the course of an infection remains poorly characterized . Here we report a comprehensive search for host sequences integrated in 21 genome populations of the baculovirus AcMNPV ( Baculoviridae ) following infection of caterpillars from two moth species . The Baculoviridae comprise large , circular dsDNA viruses infecting mainly Lepidoptera but also Hymenoptera and Diptera [25] . Most baculoviruses are transmitted as occlusion bodies ( OBs ) , i . e . the virions are protected in a protein matrix allowing the virus to remain infectious in the environment for extended periods of time [26] . AcMNPV is a multiple nucleopolyhedrovirus , meaning that each OB typically contains dozens of virions , each enclosing multiple genomes individually packaged within nucleocapsids . This morphology allows the virus to initiate infection as a highly polymorphic population [27] , and can foster the maintenance of deleterious genotypes [28] . Rather untypical for a baculovirus , AcMNPV is a generalist virus , able to infect moth species belonging to nine lepidopteran families [29] . The two moth species we used are Trichoplusia ni ( Plusiinae ) and the beat armyworm Spodoptera exigua ( Noctuinae ) , which belong to the Noctuidae family and are known to be highly susceptible to AcMNPV . These agricultural pests are found in many regions of the world , and can occur in sympatry [30] . We performed in vivo experimental infections of both T . ni and S . exigua caterpillars to generate AcMNPV populations for deep sequencing . Our population genomics approach yields the first estimate of the frequency and spectrum of host sequences that can become integrated in the genome of a large dsDNA virus .
The first AcMNPV genomic dataset we analyzed was generated by sequencing the 134-kb AcMNPV genome at 187 , 536X average depth after in vivo amplification of the virus in T . ni larvae ( G0 in S1 Fig ) . The viral population that produced this dataset was then independently passaged in ten lines of T . ni larvae and ten lines of S . exigua larvae , each line consisting of ten successive infection cycles ( G10 in S1 Fig ) . OBs recovered from the last infection cycle of each line were sequenced at between 9 , 211X and 33 , 783X average depth for the ten T . ni lines ( total depth = 145 , 386X ) and between 3 , 497X and 35 , 434X average depth for the ten S . exigua lines ( total depth = 163 , 610X ) . Viral sequencing reads were used as queries to perform Blastn searches against sequences from both moth species . Host sequences included RNAseq data corresponding to 70 , 322 T . ni contigs and 96 , 675 S . exigua contigs [13 , 14] , as well as 469 and 486 contigs from T . ni and S . exigua , respectively , that were assembled in this study using sequencing reads that did not map onto the AcMNPV genome ( see methods ) . All viral reads aligning to moth contigs were then used as queries for Blastn searches against the AcMNPV consensus genome [9] to identify chimeric reads ( i . e . sequences containing both AcMNPV and moth DNA ) , as evidence of junctions between host and viral DNA . After applying various filters to eliminate false positives , we extracted a total of 27 , 504 chimeric reads from all 21 AcMNPV genomic datasets . Chimeric reads were identified in the initial AcMNPV population from T . ni ( n = 9 , 464 ) , as well as in all ten T . ni lines ( n = 460 to 1 , 904; total = 12 , 219 ) and all ten S . exigua lines ( n = 41 to 1 , 684; total = 5 , 821 ) ( S1 Table; S1 Dataset ) . The 27 , 504 host-virus DNA junctions involved 38 T . ni and 48 S . exigua contigs ( S1 Table; S2 Dataset ) . Similarity searches and structure analyses revealed that 69 of these contigs are TEs ( 29 in T . ni and 40 in S . exigua ) belonging to both major groups of eukaryote TEs ( Table 1; S1 Table ) : retrotransposons ( three superfamilies ) and DNA transposons ( 10 superfamilies ) . The remaining 17 contigs did not show any sequence or structural similarity to any known TE or protein . However , their closest Blastn hits in the GenBank whole-genome sequence database were found in Lepidoptera , suggesting these 17 contigs indeed originate from the host genomes . The large proportion of TEs among the host contigs found to be joined to viral DNA may indicate that transposition is the main mechanism of insertion of host DNA into viral genomes . Alternatively , junctions between host TEs and viral DNA could result from technical artifact leading to chimeric reads composed of viral and contaminating host sequences . Though we verified that the amount of contaminating host DNA was very low ( if at all present ) in all our samples ( see ref [23] and S1 Text ) , we cannot totally exclude the presence of such contamination . If our samples were contaminated , and given that TEs make up the single largest fraction of eukaryote genomes [31] , technical chimeras involving mainly host TEs might not be unexpected . However , the 27 , 504 chimeric reads correspond to 7 , 049 different junctions , as defined by their location in the viral genome and in host contigs . Indeed , 1 , 412 of these 7 , 049 unique junctions are covered by more than one read ( two to 1 , 256 reads; S2 Fig ) . This strongly suggests the junctions we observe do not result from any kind of technical artifact . As duplicates generated by PCR during library construction were removed ( S1 Text ) , it seems very unlikely that a technical error would generate multiple chimeras involving exactly the same virus and host DNA sequences at the same positions . Only amplification of junctions through in vivo viral replication provides a plausible explanation for these observations , ruling out the possibility that these junctions result from technical chimeras . To further assess the biological origin of host-virus DNA junctions , we sought to characterize the molecular mechanisms involved in the integration of host sequences in AcMNPV genomes . For each host contig integrated into several distinct viral sites , we located the target insertion sites both in the viral genome and in the host sequence , and examined sequence patterns in their vicinity . We found that most inserted sequences were DNA transposons , for which the junctions with the virus genome clustered immediately before the 5’ terminal inverted repeat ( TIR ) or immediately after the 3’ TIR , as expected in the case of transposition events ( Fig 1; S3 and S7 Figs; S3 Dataset ) . In addition , the integration sites in the viral genome were generally characterized by short ( 1–5 bp ) highly conserved sequence motifs ( Fig 1; S4 Fig ) corresponding to known TE preferred insertion sites ( e . g . , TTAA for the Piggybac family , TAA for harbinger , CGNCG for transib ) . These patterns corroborate earlier findings [19–21 , 23] and indicate that many DNA transposons are indeed able to integrate into viral genomes during the course of an infection through bona fide transposition . Overall , we identified 19 , 899 host-virus junctions resulting from transposition . Counting only once all host-virus junctions covered by more than one read yields a minimum of 6 , 579 junctions resulting from independent transposition events , out of 7 , 049 . The mechanism underlying the vast majority of the host-to-virus HT detected in this study is therefore transposition . Among the remaining insertions , 434 unique host-virus junctions were deemed highly unlikely to result from transposition . Contrary to the host-virus junctions resulting from transposition , which were all located at the extremities of the 5’ and 3’ TIRs , these junctions were scattered within the host sequences . A short sequence motif of 1 to 20 bp , identical between the insertion site and the host sequence , characterized 298 of these junctions ( Fig 2 ) . The length of these microhomology motifs is significantly longer than expected by chance ( Khi2 = 4 , 523; p < 10−15 , 20 d . f . ) and argues against technical artifact as the main cause of these junctions ( as technical error favoring microhomology are highly unlikely ) . We also note that 15 of the 434 junctions are covered by more than one read ( two to 27 reads ) , indicating that some of these junctions were amplified through viral replication . The 158 remaining non-transposition junctions lacking microhomology could either have resulted from the ligation of blunt-ended sequences ( n = 87 ) , or were characterized by the presence of 1 to 2 nucleotides that apparently did not originate from either the host or viral genomes ( n = 71 , corresponding to negative microhomology lengths in Fig 2 ) . The distribution of microhomology lengths suggests that in addition to transposition , host DNA can be integrated into viral genomes via a variety of recombination events , some of which ( but not all ) rely on microhomology motifs between virus and host DNA sequences . Whether such recombination events are mediated by viral factors ( e . g . LEF-3 , AN , PCNA [32 , 33] ) , or by host-encoded DNA repair mechanisms [34] known to enhance baculovirus amplification [35] , would be worth addressing at the functional level in the future . Most of the non-transposition junctions lie within DNA transposon sequences , which may appear intriguing . We speculate that on principle , any region of the host genome could be joined to viral DNA through microhomology-mediated recombination ( including genes that may turn to be beneficial to the virus ) , provided that it contains a double stranded break . Yet , since DNA transposons have the capacity to excise themselves from the host genome , they may be among the most numerous extra chromosomal DNA fragments containing double stranded breaks ready to recombine with broken viral DNA . We then mapped the independent insertions of host sequences along the AcMNPV genomes , that is , counting only once all insertions possibly resulting from amplification through viral replication . The map ( Fig 3A; S5 Fig ) shows that integrations occur virtually everywhere in the viral genome and that all 151 viral genes are disrupted by host insertions at least once . Remarkably , the local densities of insertions of S . exigua TEs strongly correlate to those of T . ni TEs ( Fig 3B and S2 and S3 Tables; 54% of variance explained; p < 10−15 ) . This correlation is not explained by variation in sequencing depth or density of the preferred transposition motifs ( identified above ) along the viral genome ( S2 and S3 Tables ) . Two other causes may explain this correlation: ( 1 ) varying degrees of tolerance to insertions along the viral genome and ( 2 ) varying rates of transposition along the viral genome . Cause ( 1 ) implies that insertions are more likely to be replicated through viral replication in some regions than others because their impact on viral fitness would be lower . In this case , the correlation between local densities of insertions along the viral genome should be maintained or even higher when considering all insertions that possibly result from viral replication . However , the correlation almost disappears ( 0 . 1% of variance explained , S3 Table ) when all insertions ( including potentially replicated ones ) were considered . Hence , although the fitness impact of insertions may well vary along the viral genome , these variations cannot explain the strong correlation we initially observed . This leaves cause ( 2 ) as the only explanation for the correlation of insertion frequencies from TEs of the two species . In other words , TEs from the two moths , in spite of being different ( Table 1 and S1 Table ) , tend to transpose preferentially into the same regions of the AcMNPV genome irrespective of the density of target sites . This suggests that the pattern of integration may be shaped by structural properties of the viral genome . We propose that , as observed in eukaryotic genomes [36–38] , the distribution of transposition-mediated integrations along the AcMNPV genome may be influenced by accessibility of the viral genome to host TEs , which itself likely depends on the structure of AcMNPV chromatin , known to be dynamically remodeled during viral replication [39] . Taking into account the number of chimeric reads per library , the total number of reads in each library , the size of the AcMNPV genome and the minimum alignment size that can be returned by Blastn , we calculated that on average 4 . 8% viral genomes ( range 1 . 1% to 14 . 3% ) carry at least one host sequence among AcMNPV populations ( Table 1 ) . Though the number of host-to-virus HT events that generated these frequencies is likely to be high , we cannot infer it precisely because we cannot tell how many of the host-junctions covered by more than one read were amplified through viral replication . Indeed , the same TE may insert several times at any suitable viral site . Furthermore , it is possible that a number of host-virus junctions , which cannot be evaluated here , were generated through subsequent transposition of viral-borne TEs ( not coming from the host genome ) into multiple copies of the virus genome . The relatively high frequencies of AcMNPV genomes carrying host DNA fragments at any given time raise the question of whether such host sequences are inherited over viral infection cycles . We thus tested the residual presence of T . ni sequences ( inserted at G0 in S1 Fig ) in viral populations , that had subsequently been passaged 10 times in S . exigua ( G10 datasets in S1 Fig ) . Using the libraries from S . exigua G10 viruses as Blastn queries against T . ni contigs revealed 24 new insertions that were not previously found in Blastn searches against S . exigua contigs ( see S1 Text for details ) . These insertions likely involve S . exigua sequences homologous to T . ni but absent from the S . exigua contigs . None of these insertions were identical ( in terms of position in the virus genome and host contig ) to any found in the G0 virus population . The persistence of a given host DNA fragment in virus populations thus appears to be low , likely because of the deleterious effects large insertions have on the viral genome carrier . Although many new host sequences become integrated into AcMNPV genomes at each viral infection cycle , they are thus purified out of the viral population after only few infection cycles . Hence there is a high turnover of host sequences inserted into the viral genome each time the virus replicates in a host . Under natural settings , continuous host-to-virus flow of genetic material generates a significant proportion of recombinant viruses ( Table 1 ) . At the viral population scale , this represents a gene reservoir that could fuel host-virus coevolutionary arms race through co-option of a host sequence favoring the virus in a given environment . Our findings thus shed light on the first evolutionary steps underlying viral co-option of cellular genes . Under the hypothesis that viruses can act as vectors of HT of TEs , once inserted in a viral genome , viral-borne TEs should then be able to jump from the viral genome to the genome of a new host organism . To evaluate the possibility that AcMNPV can shuttle TEs between insects , we first checked whether some TEs found integrated in our AcMNPV genome datasets have retained the structural features necessary for transposition . Among the 41 contigs inserted in at least 10 different viral sites , 11 correspond to TEs for which we recovered both TIRs and that encode a putative full-length intact transposase gene ( Fig 1; S3 Fig ) . Provided that these TEs can be transcribed in a new host , they should thus be able to jump from the viral genome into the genome of this new host . However , it is important to note that the transfer would only be effective if the host survived viral infection in the first place . This is more likely to happen if the host shows resistance to the virus or if the virus harbors deleterious mutations . We then reasoned that if AcMNPV is able to act as vector of HT of TEs between its insect hosts , the TEs we found integrated in the AcMNPV genome populations might have been horizontally transferred relatively recently between insects of various susceptibility to AcMNPV . To test this hypothesis , we assessed whether some of the TEs uncovered in this study have been horizontally transferred between T . ni and/or S . exigua and other insect lineages . We used the 69 TE sequences as queries to perform Blastn searches against the 144 non-Noctuidae insect genomes available in GenBank as of March 2015 . For 21 of these TEs ( 14 S . exigua and seven T . ni TEs ) , we found highly similar copies ( >85% nucleotide identity ) in the genome of one or more other insects ( Fig 4 ) . The 21 TEs show a combination of features that are typically indicative of HT [40–42] . They have a patchy distribution in the insect phylogeny and , importantly , the between-species nucleotide identity calculated for each of these TEs is much higher ( 91% identity on average ) than synonymous nucleotide identities calculated for 11 conserved genes between the same species ( 37% identity on average , Fig 4; S4 and S5 Tables ) . We conclude that at least 21 of the 69 TEs found integrated in AcMNPV have undergone one or multiple horizontal transfers between T . ni or S . exigua and one or several other insect lineages . Nucleotide identity between some T . ni or S . exigua TEs and those uncovered in other insects is very high ( up to 99% ) , suggesting some of HT events took place very recently . Among the other insects involved in HT , we found four lepidopteran species known to be susceptible to AcMNPV infection ( Fig 4 ) [43] . Our study therefore provides further compelling support for the role of baculoviruses as potential vectors of TEs between lepidopterans [19 , 20 , 23] . Indeed , given that on average 4 . 8% of baculovirus genomes harbor a host sequence and that a caterpillar typically becomes infected by ingesting thousands of baculovirus genomes , each non-lethal infection represents an opportunity for between-host baculovirus-mediated transfer of DNA . In this study , we have shown that each time the baculovirus AcMNPV infects a caterpillar host , a large number of host TEs can transpose into its genome . Many TEs and other host sequences can also integrate into AcMNPV genomes through microhomology-mediated recombination events . Our work also demonstrates that the density of transposition events is not homogenous along the AcMNPV genome and that , while the influx of host sequences integrated into AcMNPV is continuous , each newly integrated host sequence is rapidly purged out of AcMNPV populations . Together , these observations are reminiscent of the well-known gene exchanges that take place between bacteria and bacteriophages [47] , and indicate that such exchanges may also occur on a regular basis between eukaryotes and eukaryotic viruses . Our results also raise a number of questions worth addressing in future experiments . In particular , it would be interesting to monitor the evolution of the frequency of viral replicates carrying any given host sequence across successive infection cycles , as the host DNA sequence retention time affects the likelihood of such sequence being horizontally transferred between hosts . Furthermore , since the rate of host-to-virus HT is measurable at the population level , it would be worth investigating whether this phenomenon influences the within-host replication dynamics of AcMNPV . Another exciting question is whether this phenomenon is limited to moth-AcMNPV interactions or whether it also takes place in other host-virus systems . Finally , it is noteworthy that AcMNPV and other baculoviruses are used as biopesticides and developed as vectors for several biomedical applications such as gene or vaccine delivery [48 , 49] . Our results therefore call for an evaluation of the risk of gene or TE spread through uncontrolled virus-mediated HT potentially generated by these approaches , which rely on mass production of the viruses in insect cells or in vivo .
The GenBank accession numbers of the 21 AcMNPV genomic datasets analyzed in this study are: SRS533250 , SRS534469 , SRS534534 , SRS534575 , SRS534677 , SRS534587 , SRS534590 , SRS534631 , SRS534673 , SRS536572 , SRS536571 and SRS534470 , SRS534499 , SRS534514 , SRS534536 , SRS534537 , SRS534543 , SRS534542 , SRS536937 , SRS534588 and SRS534589 [23] . They consist of 101-bp paired sequences ( reads ) , except for dataset SRS533250 , which consists in 151-bp paired reads . These datasets were produced through experimental evolution , which consisted in generating ten per os infection cycles on ten lines of T . ni and S . exigua larvae using 2500 occlusion bodies from an AcMNPV stock derived from a viral sample originally isolated from a single Alfalfa looper ( Autographa californica ) individual collected in the field . The full experiment is described in Gilbert , Chateigner [23] and in S1 Fig . The AcMNPV DNA samples used to produce the 21 sequencing datasets were all extracted using the QIAamp DNA Mini Kit ( Qiagen ) after purification of AcMNPV occlusion bodies by a percoll-sucrose gradient , Na2CO3 dissolution and enzymatic removal of host DNA [23] . Data analyses were performed in R [50] , unless another tool is mentioned . All Blastn searches were carried out under default settings . To identify host DNA sequences integrated in genomes of the AcMNPV baculovirus , we used viral reads as queries to perform Blastn searches on T . ni and S . exigua transcripts generated by Pascual , Jakubowska [51] and Chen , Zhong [52] . In addition , to recover as many insertions of host DNA as possible , we assembled non-viral DNA elements present in the viral genomic libraries . These elements may represent inserted host DNA sequences absent from ( or incomplete in ) the available transcriptomes of both moth species . We applied the following procedure on genomic libraries obtained from each moth species . We aligned all reads on the AcMNPV genome using the end-to-end mapping strategy of Bowtie 2 [53] . We used Samtools view on resulting alignment files to extract read pairs for which at least one read did not align . These unmapped reads were trimmed off low quality score bases with Trimmomatic [54] , and assembled with SOAP deNovo 2 [55] using a kmer length of 71 bases , which showed good assembly statistics compared to other lengths . We checked assembly quality by performing Blastn homology searches of assembled contigs against themselves , and found that many contigs differed only at one or both of their ends but were otherwise identical . Blastn searches of the contigs against the AcMNPV genome revealed that the contig ends that differed were similar to parts of the viral genome . We assumed that mostly similar contigs resulted from a genetic element inserted at different sites of the AcMNPV genome , and that these viral sites had been partly included into contig ends during the assembly process . We thus trimmed contigs from these viral regions , and reassembled them using the assembly feature included in Geneious 4 . 5 [56] , allowing a maximum mismatch of 10% in overlapping regions . This yielded 469 contigs for genomic libraries generated from T . ni lines and 486 contigs for S . exigua lines . These contig sequences were added to known transcriptome sequence of the corresponding moth species [51 , 52] , which we hereafter simply refer to as “transcripts” , in order to constitute the host databases for the Blastn searches designed to identify junction between moth and viral DNA . Blastn searches with default parameters [57] were carried out using the 21 AcMNPV genomic datasets as queries to identify similarities of at least 28 nucleotides ( as defined by the default settings ) between reads obtained from each viral line and the sequence database corresponding to its host . Each read showing similarities was then blasted against the AcMNPV reference genome , together with the other read of its pair ( mate ) so as to detect junctions occurring between paired reads . For a given read listed in a blast output , we retained the alignment with the best score , randomly choosing between alignments of identical scores . This selection was done separately for alignments with transcripts and for alignments with contigs , in order to help selecting between homologous contigs and transcripts ( see below ) . For a read to be considered as a junction between host and virus DNA , we imposed minimum lengths of alignment with the virus genome only , and with the host genome only , of 16 bp each ( S6 Fig ) . Furthermore , at least 95 nucleotides of the read had to align with virus and host sequences ( 130 bp for 151-bp reads ) . The overlap between these alignments was set to involve at most 20 bp and at least -2 bp . These filters excluded reads from virus regions having similarities with host contigs , in which case the region aligning to a host contig would be included in that aligning to the virus genome . To detect junctions that occurred between two paired reads ( mates ) , we selected read pairs meeting the following conditions: ( i ) one mate must have a similarity with the virus genome of at least 95 bp ( 130 bp for 151-bp reads ) and present no similarity with any host contig , and ( ii ) the other mate must have a similarity to a host contig of at least 95 bp ( 130 bp for 151-bp reads ) , and present no similarity with the virus genome . We found that the sensitivity of our approach to detect junctions between host and virus DNA was highly dependent on the quality of the assembly of the host sequences that were used as reference . Thus , it will be important that future studies dedicate substantial effort to generate a high quality and comprehensive set of host sequences in order to find all possible host-virus chimeras . We discarded all alignments ( junctions ) involving contigs or transcripts not meeting the following criteria . To ensure that a contig we assembled represented moth DNA , it had to be partly similar to a transcript of the corresponding host species , as determined by Blastn searches of contigs against host transcripts , or to align with at least one read of a pair that also aligned with a known host transcript , as determined by the Blastn search of reads against the contig and transcript databases . Because the contigs we assembled are , as expected , partly similar to some host transcripts , a chimeric read may have similarities to a contig and to a transcript ( after selecting the best alignment in each category , as explained above ) . In other words , contigs and transcripts can be candidates for the same insertions . To minimize redundancy between contigs and transcripts , any transcript sharing at least one chimeric read with a contig was discarded . We therefore retained transcripts that did not share any junction with any assembled contig . Finally , we discarded host contigs or transcripts having similarities with less than three chimeric reads ( i . e . , potentially inserted in the AcMNPV genome less than three times ) or having a cumulative alignment length of less than 75 bp with chimeric reads . For junction counts shown in S1 Table , we removed duplicates that may have resulted from PCR amplification of the same junction during library preparation ( S1 Text ) . Junctions sequenced in both directions and appearing in two overlapping paired reads were counted only once . Defining Pj as the average number of junctions per virus genome involved in the construction of a genomic library , the probability for a read from that library to cover a junction between a host sequence and the viral genome can be approximated as Pj × Lr/Lg , where Lr is the read length and Lg is the length of the virus genome . For a read to be chimeric under our criteria , a junction has to be at least 28 bp away from the read ends ( S6 Fig ) . However , the overlap between alignments at the junction ( S6 Fig ) , the mean length of which we denote as Ov , allows the junction to be slightly closer from the read end and to yield a 28-bp region of sequence similarity detectable by blast , so that the probability for a read to be chimeric is Pj ×Lr−56+OvLg . Since this probability can be approximated as the ratio of the number of chimeric reads Nc over the number of viral reads N of a sequence library , we obtain Pj≅NcN × LgLr−56+Ov . N was estimated by running samtools view [58] on alignment files obtained by mapping reads from each virus line on the AcMNPV consensus genome , using the local sensitive settings of Bowtie 2 [53] . Nc includes technical duplicates ( PCR duplicates and overlapping paired reads , see above ) because they contribute to N as much as they do to Nc . We derived the proportion of virus genomes carrying at least one host DNA fragment by assuming that the number of inserted fragments per virus genome follows a Poisson distribution of mean Pj/2 , as one insertion of host DNA into the circular AcMNPV genome should yield 2 junctions . For simplification , we hereafter refer to contigs assembled in this study and to previously assembled host transcripts as “contigs” . We identified each junction producing a chimeric read by the offset it involves between the virus genome coordinates and the host contig coordinates ( S1 Text ) . Reads having the same offset , involving the same viral DNA strand , host contig , and coming from the same genomic library in the case of S . exigua lines ( which do not share a ancestor on this host ) were considered likely to come from viral amplification of the same original insertion of host DNA . In the following analyses , we selected only one chimeric read per original junction , favoring the read with best alignment score on the host contig . These reads were mapped onto their corresponding contig by inserting gaps of appropriate length before their sequence , based on alignment coordinates reported by blast , to produce multifasta alignment files . Visualization of these files in Geneious [56] and BioEdit [59] ( example shown in S7 Fig ) showed that junctions clustered at one or two positions in a contig likely representing the end ( s ) of a TE . For many contigs , this was further supported by the presence of terminal inverted repeats ( TIRs ) , which are typical of class II DNA transposons , the presence of long terminal repeats ( LTRs ) , which are typical of LTR retrotransposons and by similarities to known TE protein motifs returned by blastx on the GenBank non-redundant protein database . For each cluster of at least 10 junctions , which likely represent insertions of the same TE end , we analyzed sequence conservation at insertion sites in the virus genome . This was done by computing insertion coordinates based on the previously obtained offset ( S1 Text ) , and by building sequence conservation logos [60–62] of 30-bp sequences around insertion sites . Sequence logos are provided in S4 Fig . Some junctions did not form clusters ( according to our criteria defined in S1 Text ) and were scattered along host contigs ( example shown in S7 Fig ) , suggesting that different fragments of these contigs were inserted . This concerned 434 junctions , most of which presented similarities between host and virus sequences at insertion points ( Fig 2 , yielding to the overlap shown in S6 Fig ) . To check whether these similarities were overall longer than expected by chance , we extracted from each chimeric read resulting from this type of junction the last 20 bp that aligned to the host contig ( next to the junction point ) , and computed the lengths of similarities this 20-bp sequence had with 20 random 20-bp regions of the AcMNPV consensus genome . This allowed comparing the distribution of expected and observed identity lengths with a Khi-square test . We explained the number of junctions in 1500-bp windows of the AcMNPV genome , combining all virus lines from S . exigua , with a generalized linear model including three covariates: the average sequencing depth in that window , the number of common TE targets found in S . exigua lines , and the number of junctions found in virus lines from T . ni , without considering interactions between terms . Sequencing depth was estimated by running samtools mpileup [58] on mapping files obtained previously . We modified mpileup to allow greater depth than 8000 . We counted the following frequent targets of S . exigua TEs , based on the logos we established previously ( S2 Fig ) : “TTAA” ( for piggybac TEs ) , “TTA” , “TAA” ( for Harbinger TEs ) , and “TA” ( for Mariner TEs ) . A Poisson distribution was assumed for the number of junctions per genome window . We selected the best model on the basis of corrected Akaike Information Criterion ( AICc ) returned by the dredge function of the R package MuMIn [63] ( S2 Table ) , and we submitted it to an analysis of deviance ( S3 Table ) . We fitted this model twice: considering all junctions ( including viral replicates ) , and only independent junctions based on their identifiers . In the latter case , the most likely model only included the number of junctions per window in T . ni as a covariate ( S2 Table ) . Sequencing depth of host contigs ( Fig 1; S3 Fig ) was computed by using alignments coordinates from results of Blastn search of reads against host databases ( see above ) , using the same criteria to select a single alignment for reads having similarities with several contigs/transcripts . Depth was averaged over 20-bp sliding windows overlapping by 10 bp . We assessed whether T . ni and S . exigua TEs found integrated in the AcMNPV genome underwent HT between insects . We used the T . ni and S . exigua TEs we identified as queries to perform Blastn searches against the 144 non-Noctuidae insect whole genome sequences available in GenBank as of March 17th 2015 . We identified candidate HT events when a T . ni or S . exigua TE aligned to a sequence from another insect genome with at least 85% nucleotide similarity over at least 100 bp . To assess the level of neutral genetic distance expected under vertical inheritance between T . ni/S . exigua and all insect species in which we found Blastn hits meeting the above criteria , we calculated synonymous distances for 11 conserved genes between T . ni/S . exigua and those insect species using the non-corrected Nei-Gojobori method in MEGA 6 [64] , following Gilbert , Chateigner ( 23 ) . Overall we calculated 143 pairwise synonymous gene distances between S . exigua and 13 other insect species and 55 pairwise synonymous gene distances between T . ni and five other insect species . | While gene exchange is known to occur between viruses and their hosts , this phenomenon has never been studied at the level of the viral population . Here we report that each time a virus from the Baculoviridae family infects a moth , a large number ( dozens to hundreds ) and high diversity of moth DNA sequences ( 86 different sequences ) can integrate into replicating viral genomes . These findings show that viral populations carry a measurable load of host DNA sequences , further supporting the role of viruses as vectors of horizontal transfer of DNA between insect species . The potential uncontrolled gene spread associated with the use of viruses produced in insect cells as gene delivery vectors and/or biopesticides should therefore be evaluated . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"sequencing",
"techniques",
"invertebrates",
"moths",
"and",
"butterflies",
"microbiology",
"animals",
"invertebrate",
"genomics",
"genomic",
"databases",
"viral",
"genome",
"genome",
"analysis",
"sequence",
"motif",
"analysis",
"molecular",
"biology",
"techniques",
"microbial",
"genomics",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"viral",
"genomics",
"genomic",
"libraries",
"sequence",
"alignment",
"biological",
"databases",
"molecular",
"biology",
"insects",
"animal",
"genomics",
"arthropoda",
"virology",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"computational",
"biology",
"organisms"
]
| 2016 | Continuous Influx of Genetic Material from Host to Virus Populations |
The mechanisms by which the gut luminal environment is disturbed by the immune system to foster pathogenic bacterial growth and survival remain incompletely understood . Here , we show that STAT2 dependent type I IFN signaling contributes to the inflammatory environment by disrupting hypoxia enabling the pathogenic S . Typhimurium to outgrow the microbiota . Stat2-/- mice infected with S . Typhimurium exhibited impaired type I IFN induced transcriptional responses in cecal tissue and reduced bacterial burden in the intestinal lumen compared to infected wild-type mice . Although inflammatory pathology was similar between wild-type and Stat2-/- mice , we observed decreased hypoxia in the gut tissue of Stat2-/- mice . Neutrophil numbers were similar in wild-type and Stat2-/- mice , yet Stat2-/- mice showed reduced levels of myeloperoxidase activity . In vitro , the neutrophils from Stat2-/- mice produced lower levels of superoxide anion upon stimulation with the bacterial ligand N-formylmethionyl-leucyl-phenylalanine ( fMLP ) in the presence of IFNα compared to neutrophils from wild-type mice , indicating that the neutrophils were less functional in Stat2-/- mice . Cytochrome bd-II oxidase-mediated respiration enhances S . Typhimurium fitness in wild-type mice , while in Stat2-/- deficiency , this respiratory pathway did not provide a fitness advantage . Furthermore , luminal expansion of S . Typhimurium in wild-type mice was blunted in Stat2-/- mice . Compared to wild-type mice which exhibited a significant perturbation in Bacteroidetes abundance , Stat2-/- mice exhibited significantly less perturbation and higher levels of Bacteroidetes upon S . Typhimurium infection . Our results highlight STAT2 dependent type I IFN mediated inflammation in the gut as a novel mechanism promoting luminal expansion of S . Typhimurium .
A healthy gastrointestinal microbiota is characterized by the dominance of obligate anaerobic members of the phyla Bacteroidetes and Firmicutes . The expansion of facultative anaerobic Enterobacteriaceae ( phylum Proteobacteria ) is considered a microbial signature for gut inflammation and dysbiosis [1 , 2] . This signature is observed in severe human intestinal diseases including inflammatory bowel disease ( IBD ) , [3–5] colorectal cancer [6] and necrotizing enterocolitis [7] . Several mechanisms by which the enteric pathogen , Salmonella enterica serovar Typhimurium , capitalizes on multiple processes induced by inflammation and outcompete the commensal have been described . Infection with S . Typhimurium starts with the invasion of intestinal epithelial cells using its type III secretion system ( T3SS-1 ) [8] . After crossing the intestinal barrier , the bacterium is rapidly recognized by Pattern Recognition Receptors ( PRRs ) , such as Toll-like receptors ( TLRs ) and Nod-like receptors ( NLRs ) , and is internalized by macrophages or dendritic cells . In macrophages , S . Typhimurium survives using its T3SS-2 [9] . Epithelial invasion , recognition of Pathogen-Associated Molecular Patterns ( PAMPs ) and macrophage survival leads to the production of chemokines and cytokines triggering an inflammatory environment and acute colitis [10–12] . In the lumen , S . Typhimurium employs mechanisms to utilize unique respiratory electron acceptors ( e . g . tetrathionate and nitrate ) which are generated as byproducts of the inflammatory burst . Most commensal members of the microbiota are unable to metabolize nitrate and tetrathionate [13 , 14] . As a result , S . Typhimurium outcompetes the healthy microbiota enabling its luminal expansion and eventually facilitating the transmission to subsequent hosts [13–16] . Although S . Typhimurium succeeds in expanding its luminal population during inflammation leading to a decline in the commensal microbiota , the coordinated actions of multiple immune cell defense pathways mediate the clearance of the pathogen . Activation of the Interferon ( IFN ) signaling pathway is critical for successful host defense against many infections . Type II IFN ( IFN γ ) plays a central role in generating inflammatory responses to clear S . Typhimurium [17–20] . However , the role of a closely related pathway involving the actions of type I IFNs ( IFNs α and β ) is less clear . Type I IFN signaling is well-documented as essential for mounting antiviral responses . Pre-exposure of cells to type I IFNs induces an antiviral state by blocking viral replication [21 , 22] . It has recently become evident that activation of this pathway also plays a pivotal role during bacterial infections by acting directly or indirectly on many immune cell types including NK cells , T cells , B cells , Dendritic Cells ( DCs ) , neutrophils and phagocytic cells . Depending on the bacterial agent , the role of type I IFNs exert seemingly opposing roles . For instance , while type I IFNs restrict the growth of Legionella pneumophila or Streptococcal species [23–27] , activation of the same pathway impairs the clearance of intracellular Mycobacterium tuberculosis leading to tuberculosis [28 , 29] . Recent studies highlighted the role of type I IFN signaling during systemic infection with S . Typhimurium . Mice deficient in type I IFN receptor ( IFNAR ) , or IFN β exhibit greater resistance to S . Typhimurium [30] . Furthermore , type I IFNs are critical for inflammasome formation , caspase activation , and inflammatory cell death following infection with S . Typhimurium [31–34] . The role of this pathway during intestinal bacterial induced inflammation and the subsequent impact on the luminal bacterial population remains unclear . IFNAR activation by type I IFNs ( IFNs α and β ) not only leads to the transcription of type I IFN stimulated genes ( ISGs ) induced by ISGF3 , the heterotrimeric transcriptional complex composed of STAT1/STAT2/IRF9 , but also by inflammatory gene activation via the formation of STAT1 homodimers . As STAT1 homodimers can also be activated by IFN γ , earlier studies that used Ifnar-/- or Stat1-/- mice did not clearly differentiate the contribution of each IFN pathway to driving inflammation ( Fig 1 ) . Here we used Stat2-/- mice , which causes the genetic ablation of type I IFN signaling , in combination with the streptomycin pretreated mouse model to pinpoint the role of type I IFNs in host response to Salmonella infection . Overall , we conclude that STAT2-driven type I IFN response leads to the transmigration of functional neutrophils into the lumen creating a microaerophilic environment , which enables the pathogen to outgrow the microbiota .
Type I IFNs released during bacterial infections may affect many arms of the immune response including inhibition of bacterial invasion , amplification of the immune response and production of antimicrobial genes . To investigate the possible role of a STAT2-dependent type I IFN signaling pathway during S . Typhimurium induced intestinal infection , wild-type C57BL/6 , Stat1-/- ( deficient in both IFN-α/β and IFN-γ signaling ) , and Stat2-/- ( deficient only in IFN-α/β signaling ) mice were orally infected with 109 CFU of S . Typhimurium following streptomycin pretreatment . All strains eventually succumbed to S . Typhimurium infection , but Stat2-/- mice survived significantly longer than wild-type and Stat1-/- mice ( p = 0 . 0026 ) ( Fig 2 ) . This finding is notable because our previous study [35] reported increased mortality with Stat2-/- mice during LPS-induced sepsis suggesting that type I IFNs play different roles when compared between mucosal and systemic sites during infection . To determine the role of STAT2 during S . Typhimurium infection , we first evaluated intestinal immune responses by analyzing gene expression by qPCR in the cecum of mice at 48 hours post infection , a time point where no animal death was observed and found to be optimal for investigating inflammatory responses [36] . When we examined the expression of genes that have previously been identified to be dependent on STAT2 [37] , we found that there were significantly lower transcript levels of Irf7 , Isg15 , Oas1b , Rsad1 , and IrgM1 in the cecum of infected Stat2-/- mice when compared to cecum of infected wild-type mice ( Fig 3 ) . We found no significant differences in genes known to be regulated by IFN γ and the IFNGR such as Cxcl10 ( Fig 3 ) . Furthermore , no differences in the transcription levels of genes previously shown to be important for S . Typhimurium infections including Tnfα , Il6 , Ifnγ , and Mcp1 were observed between wild-type and Stat2-/- mice ( Fig 4 ) . This result was further confirmed when we analyzed the systemic cytokine responses in the serum using a cytometric bead assay . We did not observe a significant difference in the serum levels of TNFα , IFNγ , MCP1 ( also known as CCL2 ) , IL-12 , IL-6 or IL-10 between wild-type and Stat2-/- infected mice ( S1 Fig ) . Overall , these results show that type I IFN signaling is distinctively blocked in Stat2-/- mice as classical inflammatory gene expression was unaffected by this deficiency . When we investigated the bacterial burdens at 48 hours post infection , the time point where we observed changes in immune responses , we found that there were significantly fewer bacteria in the cecum and colon contents of Stat2-/- mice compared to wild-type mice ( Fig 5 ) . No differences were observed in bacterial numbers in mesenteric lymph nodes ( MLN ) . Although there was a trend towards lower numbers in spleen and liver at this time point ( Fig 5 ) , it was not statistically significant . The fact that a deficiency in STAT2 signaling leads to decreased bacterial burden specifically in the lumen suggests a role for a STAT2-induced inflammatory environment in S . Typhimurium expansion . In response to infection with S . Typhimurium , neutrophils migrate into the tissue as well as the lumen [38 , 39] . Studies using different pathogens have suggested that type I IFNs not only mediate the migration of neutrophils into the infection site but also enhance their function [40 , 41] . To determine whether there was a defect in neutrophil migration as well as pathology , cecal tissue samples were fixed and stained with H&E . No differences were observed in overall histopathology between wild-type and Stat2-/- mice infected with wild-type S . Typhimurium at 48 hours ( S2 Fig ) . Neutrophil numbers ( PMN/field ) were similar between wild-type and Stat2-/- mice infected with wild-type S . Typhimurium ( Fig 6A ) . No differences in neutrophil abundance were noted when comparing uninfected wild-type and Stat2-/- ( S3A Fig ) . However , when we quantified levels of myeloperoxidase ( MPO ) , a neutrophil marker , we surprisingly found that there was less MPO in the cecal tissue of S . Typhimurium-infected Stat2-/- mice than infected wild-type mice ( Fig 6B ) . This result was confirmed using immunohistochemistry with an MPO-specific antibody ( Fig 6C ) . These results suggested that although the presence of type I IFNs does not affect the transmigration of neutrophils into the infection site , they somehow alter the function of these immune cells . This finding is somewhat surprising because neutrophils play a major role in clearing S . Typhimurium . The fact that there were more bacteria in the presence of neutrophils indicated to us that a novel mechanism allows this pathogen to thrive in the presence of neutrophils . To determine if intestinal oxygenation caused the difference in bacterial burdens in the intestines of wild-type versus the Stat2-/- mice , we infected mice with the S . Typhimurium cyxA mutant . The cyxAB operon encodes a cytochrome bd-II oxidase enzyme that facilitates growth of S . Typhimurium under oxygen-limiting conditions [42–45] . CyxA is essential for S . Typhimurium survival in the post-antibiotic treatment model [45] . Mice were streptomycin pretreated and then orally administered a 1:1 mixture of wild-type S . Typhimurium and cyxA mutant . Four days post infection colon contents were collected for bacterial enumeration by determining colony forming units ( CFU ) , and the competitive index ( CI ) was determined by dividing the output ratio ( wild-type CFU/cyxA CFU ) in the colonic contents of mice by the input ratio ( wild-type CFU/cyxA CFU ) . Wild-type S . Typhimurium exhibited a fitness advantage over the cyxA mutant in wild-type mice ( higher numbers of wild-type bacteria recovered ) , consistent with previous findings [45]; however , the cyxA gene provided no advantage in the Stat2-/- mice ( both strains were recovered at same numbers ) ( Fig 6D ) . There was no observable phenotype in systemic sites such as the liver where there was no fitness advantage conferred by the cyxA mutant in either the wild-type or STAT2-/- mice ( Fig 6E ) . These data suggested that oxygenation in the intestine of wild-type mice is different from that of Stat2-/- mice . This was confirmed using pimonidazole ( PMDZ ) , a marker of hypoxia . Mice were injected intraperitoneally with PMDZ ( Chemicon; 2 . 0 mg/20 g body weight in 100 μl PBS ) 1 hour prior to euthanasia , and PMDZ was detected in tissue sections by immunohistochemistry . It was previously reported that hypoxia decreases in the intestine during S . Typhimurium infection [45] . We determined that the intestinal environment in Stat2-/- mice was more hypoxic than in wild-type mice as shown by higher levels of pimonidazole staining ( red ) ( Fig 6F ) . Both wild-type and Stat2-/- mice showed comparable hypoxia staining without infection ( S3B Fig ) . Obligate anaerobes of the healthy gut microbiota were previously reported to become depleted from the microbiota at later stages of S . Typhimurium infection in streptomycin-treated mice through a neutrophil-dependent mechanism [46] . To determine if STAT2 signaling led to changes in the microbiota , we analyzed the phylogenetic composition of the intestinal microbial communities using 16S rRNA profiling ( S4A Fig ) . We observed a drastic reduction in the relative abundance of Bacteroidetes phylum with approximately 10% remaining in wild-type mice infected with S . Typhimurium as opposed to 60% in uninfected wild type mice . No significant shifts were detected in the abundance of Bacteroidetes in Stat2-/- infected mice when compared to that of uninfected wild-type and Stat2-/- control mice ( S4B Fig ) . Conversely to the CFU recovered from feces of wild-type and Stat2-/- mice upon S . Typhimurium infection ( Fig 5 ) , there was a significantly higher relative abundance of Proteobacteria observed in the wild-type infected mice with an average of 70% than in the Stat2-/- infected mice ( 30%; S4C Fig ) . To verify that the experimental changes we observed between wild-type and Stat2-/- mice was not due to differences in the overall microbiome content of the two strains of mice that arose because they were housed separately , we co-housed the mice starting at the day of weaning for 5 weeks . The co-housed wild-type and Stat2-/- mice were infected with S . Typhimurium following streptomycin pre-treatment . Forty-eight hours post infection , fecal and cecal contents were collected and the phylogenetic composition of the microbial communities at the phylum level was determined using 16S rRNA profiling ( Fig 7A and 7B ) . Uninfected mice were also included as controls . Infection of wild-type mice with S . Typhimurium led to a reduction in the relative abundance of the Bacteroidetes phylum . The relative abundance of Bacteroidetes in Stat2-/- infected mice remained comparable to that of uninfected control mice ( Fig 7C ) . It’s important to emphasize that the striking reduction in Bacteroidetes abundance we detected earlier in non-cohoused infected and non-infected wild type mice were no longer observed under co-housing conditions . Nevertheless , differences between infected wild type and Stat2-/- mice remained unchanged . Moreover , relative abundance of Proteobacteria was increased in the wild-type infected mice and this expansion was at a lower level in Stat2-/- infected mice ( Fig 7D ) . Detailed microbial analysis revealed that the members of Proteobacteria that expanded in their relative abundance in infected wild-type mice belonged to the Salmonella genus ( S5 Fig ) . This result was validated by directly enumerating S . Typhimurium in the colon contents of co-housed infected wild-type and Stat2-/- mice ( Fig 7E ) . Overall , the results obtained in both experiments using co-housed and non-cohoused mice demonstrated that STAT2 enabled the expansion of S . Typhimurium . These data also strongly indicate that post-infection , Stat2-/- mice retained a protective microbiome against pathogenic bacteria . To also confirm the previous findings on neutrophils ( Fig 6 ) , we performed a semi-quantitative analysis of overall pathology and quantified neutrophil abundance . There were no differences in the overall pathology ( Fig 8A ) and the neutrophil numbers between the S . Typhimurium infected wild-type and Stat2-/- mice ( Fig 8B ) . The cecal neutrophils were also quantified using flow cytometry . Following the identification of live cells , neutrophils were identified as CD45+ , CD3- , NK1 . 1- , B220- , Ly6G+ cells using the gating strategy described in S6 Fig . While there was an increase in percentage of neutrophils in wild-type mice infected S . Typhimurium compared to uninfected wild-type mice , there were no significant difference observed in the percentage of neutrophils when comparing S . Typhimurium infected wild-type and with S . Typhimurium infected Stat2-/- mice ( Fig 8C ) . As we did not observe any differences between numbers of neutrophils transmigrating into the infection site between wild-type and Stat2-/- mice but there was a difference in MPO levels in the colon contents of these mice ( Fig 6C ) , we next determined whether Stat2-/- neutrophils were functional . Bone marrow neutrophils from wild-type and Stat2-/- mice were stimulated with the Gram-negative bacterial ligand N-formylmethionyl-leucyl-phenylalanine ( fMLP ) in the absence or presence of IFNα . Superoxide anion generation was then measured . There were no differences in superoxide anion generation between the neutrophils of wild-type and Stat2-/- mice upon stimulation with fMLP . However , when the cells were pre-treated with a type I IFN , IFNα , there were reduced levels of superoxide anion generation in the neutrophils isolated from Stat2-/- mice compared to wild-type mice ( Fig 8E )
The immune system deploys multiple mechanisms to eradicate invading microbes and infections . Induction of type I IFNs is a critical mechanism that the immune system exploits to fight viral infections . Type I IFNs ( IFNs α and β ) induce antiviral responses by binding to their cognate receptor IFNAR ubiquitously expressed on many cell types . The transcription factor STAT2 takes center stage in the type I IFN response as it is essential to mediate an antiviral state that helps the host clear a viral infection [47] . Research over the past few years has suggested that type I IFNs are intricate players during bacterial infections . Although type I IFN responses mounted against viral infections provide a common anti-viral state among a broad range of viral pathogens , the type I IFN response generated against bacteria varies based on the specific bacterial pathogen . Recently , it was reported that Ifnβ−/− mice exhibit greater resistance to oral S . Typhimurium infection and a slower spread of S . Typhimurium to distal sterile sites [30] . These results are consistent with our findings using Stat2-/- mice ( Fig 2 ) . Nevertheless , the previous study did not use streptomycin pre-treatment to induce colitis during infection , which models more accurately the course of S . Typhimurium infection . Hence the role of type I IFNs during gut inflammation and dysbiosis has remained unclear . Several studies have emerged showing that not all type I IFN responses involve the classical ISGF3 complex . STAT2 homodimers have been shown to bind IRF9 and activate ISG expression of antiviral genes in the absence of STAT1 [37 , 48] . The expression of a subset of ISGs stimulated by STAT2 homodimers/IRF9 exhibits a delayed kinetics compared to the classical ISGF3 , however , this is sufficiently robust to evoke an innate response [49] . These observations together with our own findings indicate that S . Typhimurium exploits the type I IFN pathway by relying on STAT2 and potentially in the absence of STAT1 . S . Typhimurium successfully establishes an infection with the coordinated actions of its two distinct populations; the first invades the tissue and increases inflammation while the second luminal population counter intuitively benefits from the generation of host derived nitrate and oxygen [14 , 15 , 50] . The regulatory host signaling pathways that control the availability of these electron acceptors are not known . Here , we determined that type I IFN pathway is activated during S . Typhimurium infection and leads to oxygenation of the gut mucosa allowing the pathogen to respire and expand its luminal population . As we observed blunted expression of type I IFN stimulated genes ( Fig 3 ) , similar numbers of neutrophils in the cecal mucosa but lower levels of MPO ( Fig 6A and 6B ) , a neutrophil activation marker , in the cecum of Stat2-/- mice infected with S . Typhimurium , these results suggest that type I IFNs do not effect the migration of neutrophils to the site of infection but may effect the antimicrobial activity of these cells . It was previously established that upon activation , neutrophils release reactive oxygen species as antimicrobial measures . The release of reactive oxygen species also contributes to oxygenation of the lumen , and superoxide dismutases encoded by Salmonella allow the bacteria to detoxify the oxygen radicals promoting bacterial survival [51] . The competition experiments between wild-type S . Typhimurium and cyxA mutant as well as hypoxia staining confirmed that oxygenation of the gut lumen of Stat2-/- mice was lower compared to that of wild-type mice . Furthermore , our in vitro experiments confirmed that neutrophils from the Stat2-/- mice were blunted in their ability to generate superoxide anion . Overall , these results suggest that in response to S . Typhimurium , neutrophils invade the gut lumen and contribute to the oxygenation of the gut via a type I IFN mediated mechanism . In return , the professional pathogen S . Typhimurium takes advantage of this mechanism and expands its population . One of the many benefits of the gut microbiota to the host is to limit the expansion of enteric pathogens . Gut microbiota provides metabolites such as butyrate that fuels colonocyte metabolism resulting in the consumption of oxygen , thereby rendering the lumen hypoxic . It was recently shown that the epithelial PPAR-γ-signaling pathway limits oxygenation of the gut epithelium in the presence of butyrate , which in turn limits the expansion of S . Typhimurium [50] . In addition to the neutrophils , we do not know whether STAT2 signaling may also have a direct effect on colonocyte metabolism ( Fig 6D ) , which impacts the bioavailability of oxygen in the gut lumen . Previous studies have shown that the abundance of dominant microbial phyla , Bacteroidetes and Clostridia can directly be affected by drastic changes in the luminal environment during enteric infections [52–54] . The depletion of Clostridia from the microbiota at later stages of S . Typhimurium infection in streptomycin-treated mice through a neutrophil-dependent mechanism was reported [46] . Our results demonstrate that a STAT2 mediated type I IFN response triggered during S . Typhimurium infection directly affects Bacteroidetes phyla in the gut . Our study highlights the importance of STAT2 signaling in neutrophils during Salmonella infection . To date , most of what has been described for STAT2 signaling in pathogenic infections was centered on immune cells , such as macrophages and dendritic cells . In models of viral infection , STAT2 signaling is exploited by measles virus and choriomeningitis virus to interfere with dendritic cell ( DC ) development and expansion [55] . Furthermore , STAT2 signaling in macrophages is critical to activate a transcriptional response and control early dengue virus replication [37 , 56] . We speculate that STAT2 signaling in colonocytes during Salmonella infection is equally important as in neutrophils for crosstalk and the release of chemokines for the recruitment and activation of neutrophils and macrophages . Future studies involving a more detailed analysis are warranted to delineate the far-reaching effects of type I IFN signaling on both the microbiota and oxygenation by colonocytes and neutrophils .
Salmonella enterica serovar Typhimurium strain IR715 , a fully virulent , spontaneous nalidixic acid resistant derivative of strain ATCC 14028 , was grown in Luria-Bertani broth ( LB ) supplemented with 50 μg/ml nalidixic acid at 37°C [57] . Salmonella Typhimurium IR715 cyxA [45] , generously provided by Andreas Baumler , was supplemented with 100 μg/ml carbenicillin LB broth and was grown at 37°C . Eight- to ten-week-old female C57BL/6 ( wild-type ) mice were age and sex matched to mice deficient in STAT1 ( kindly provided by Dr . David Levy , NYU ) or STAT2 ( generously provided by Dr . Christian Schindler on the SvJ background that we backcrossed 10 generations onto the B6 genetic background ) . All mice were bred at the animal facility of the Lewis Katz School of Medicine at Temple University . All mice were streptomycin treated prior to bacterial infection . Mice were monitored twice daily after infection . Humane terminal endpoints included inability to ambulate and/or labored breathing . Briefly , mice were inoculated intragastrically with 20 mg of streptomycin ( 0 . 1 ml of a 200 mg/ml solution in water ) 24 hours prior to bacterial infection . Bacteria were grown with shaking in LB broth containing nalidixic acid ( 50 μg/ml ) at 37°C overnight . For infection , groups of 3 to 5 mice were inoculated intragastrically with either 0 . 1 ml of sterile LB broth ( mock infection ) or 109 CFU of S . Typhimurium . Mice were sacrificed at indicated time points after infection . To determine the number of viable S . Typhimurium , samples of cecum ( proximal section ) , liver , spleen , mesenteric lymph nodes , and colon contents were collected from each mouse and homogenized in 5 ml PBS . 10-fold serial dilutions were plated on LB agar plates containing nalidixic acid ( 50 μg/ml ) . The tip of the cecum was collected for histopathological analysis . The center section of the cecum was immediately snap-frozen in liquid nitrogen and stored at −80°C for RNA isolation . All animal experiments were repeated at least three times with identical results . To determine to role of luminal oxygenation in bacterial survival , 24 hours prior to inoculation 6 to 8 week-old age matched wild-type C57BL/6 and Stat2-/- mice were orally gavaged 0 . 1 ml of a 200mg/ml streptomycin solution . Mice were orally infected with 108 bacteria in a 1:1 ratio of S . Typhimurium IR715:cyxA . Four days after infection mice were euthanized and colon contents , cecum and liver were collected to determine the CFU of IR715 and cyxA mutant . The cecum was snap frozen in liquid nitrogen and stored at -80°C for MPO ELISA , and sections of the colon were collected for histopathological analysis . Organs for bacterial enumeration were homogenized as mentioned above and plated on selective media using 10-fold serial dilutions . The competitive index ( CI ) was calculated as the ratio of recovered bacterial strains ( output ratio ) divided by the ratio present in the inoculum ( input ratio ) . All animal experiments were at least repeated three times with identical results . RNA was extracted from snap-frozen tissues or tissue culture cells using 1 ml TriReagent ( Molecular Research Center , TR118 ) according to the manufacturer's protocol . RNA was then treated with DNase according to the manufacturer’s protocol ( Ambion , AM1906 ) . Reverse transcription of total RNA ( 1 μg ) was performed in 25 μl volume according to manufacturer's instructions using the TaqMan Reverse Transcription Kit ( Invitrogen , N8080234 ) . Real-time PCR was performed using the SYBR green ( Applied Biosystems , 4309155 ) or TaqMan ( Applied Biosystems ) according to the manufacturer's instructions . Real-time PCR was performed for each cDNA sample ( 5 μl per reaction ) in duplicate using the Step One Plus real-time PCR system ( Applied Biosystems ) . The primers sequences are listed in Table 1 . Results were analyzed using the comparative ΔCT method . Data was normalized to Gapdh or β-actin for SybrGreen or TaqMan reagents , respectively . Fold increases in gene expression in infected or mock-infected Stat2-/- mice were calculated relative to the average level of the respective cytokine in the mock-infected wild-type mice . Hypoxia studies were performed as described by the manufacturer’s instructions ( Hypoxyprobe-1 Plus Kit , Chemicon , Temecula , CA , USA ) [45] . One hour prior to euthanasia , wild-type and Stat2-/- infected mice were injected with 100 mg/kg of PMDZ diluted in DMSO . After euthanasia , colon samples were collected and fixed with 10% formalin . Unstained paraffin embedded tissue samples were probed with 1:50 FITC-conjugated IgG1 mouse monoclonal antibody clone 4 . 3 . 11 . 3 ( Hypoxyprobe , Inc . ) , and stained with 1:150 Cy3 conjugated AffniPure Goat Anti Mouse IgG ( H+L ) ( Jackson ImmunoResearch , 115-165-0003 ) . Briefly , tissue sections were incubated at 50°C for 10 minutes and then deparaffinized by washing for 10 minutes with xylene 2x , 3 minutes with 95% ethanol 2x , 3 minutes with 80% ethanol 1x , and then rehydrated by washing with 70% ethanol 1x . The antigens were retrieved by incubating sections with 20μg/ml Proteinase K ( Fisher , BP1700-100 ) in TE buffer ( 10mM Tris , 1mM EDTA , pH 8 . 0 ) for 15 min at 37°C in a humidified chamber . The slides were washed with PBS for 10 minutes and then blocked for 45 minutes with blocking buffer . Samples were incubated with the primary antibody 1: 50 FITC-conjugated IgG1 mouse monoclonal antibody clone 4 . 3 . 11 . 3 over night at 4°C in a humidified chamber . After PBS washing ( 5 minutes , trice ) , each slide was incubated with the secondary antibody 1:150 Cy3 conjugated AffniPure Goat Anti Mouse IgG ( H+L ) at room temperature for 90 minutes in a humidified chamber . DAPI ( Invitrogen , P21490 ) was used as a counter-stain ( 1μg/ml , incubated at room temperature for 5 minutes in the dark ) . Slides mounted with Vectashield ( Vector Labs , H-1000 ) and were visualized using an Olympus BX60 Fluorescent Microscope with Spot Insight2 camera at 10x magnification . MPO activity in the cecal tissue was determined as previously described [58] . Snap frozen cecum samples were lysed by homogenizing in 0 . 5% HETAB ( hexadecyltrimethyl ammonium bromide ) in 50mM KPi ( phosphate buffer ) at a ratio of 0 . 1g sample per 1ml buffer . Master mixes were prepared by mixing 10 μL homogenized sample with 3 μL o-dianisidine hydrochloride ( 20mg/ml stock ) , 3 μL 20 mM hydrogen peroxide and 284μL 50mM KPi . Ten-fold serial dilutions of the MPO standard 100UG ( Millipore , 475911 ) were prepared in the same fashion as the samples , with the top standard being 125μg/ml . Samples were plated in clear 96 well plate and incubated at 37°C for 10 minutes , taking absorbance measurements at 460 nm every 2 minutes for 10 minutes using a Flex Station , Molecular Devices plate reader . The reaction was halted by adding 3 μL of 30% NaN3 to each well and a final absorbance reading at 460nm was taken . The concentration of MPO was calculated using the absorbance values obtained from the standard curve . To visualize the presence of MPO within the cecum of wild-type C57BL/6 and Stat2-/- mice , unstained paraffin embedded tissue sections were heated at 55–60°C for 30 minutes . The tissue was deparaffinized by washing in xylene 2x for 5 minutes , absolute ethanol 3x for 3 minutes , 95% ethanol 1x for 3 minutes , 90% ethanol 1x for 3 minutes , 70% ethanol 1x for 3 minutes . Antigens were retrieved by boiling slides in sodium citrate buffer ( 10mM Sodium Citrate , 0 . 05% Tween 20 , pH 6 . 0 ) , for 10 minutes , and allowing to cool to room temperature for 20 minutes . The samples were washed in Tris Buffered Saline , TBS ( 0 . 5M Tris Base , 9% NaCl ) 1x for 5 minutes . The tissue was blocked in TBS supplemented with 3% BSA for 30 minutes and then incubated with 1:200 MPO Heavy Chain ( L-20 ) , ( Santa Cruz sc-16129 ) in TBS supplemented with 3% BSA for two hours at room temperature . The samples were washed 3x for 5 minutes each in TBS supplemented with 3% BSA . The samples were then incubated for 40 minutes at room temperature with 1:1000 Rabbit anti-Goat ( H+L ) Super Clonal Secondary antibody , Alexa Fluor 488 conjugated ( Fisher A27012 ) . The slides were washed with 3x for 5 minutes each in TBS supplemented with 3% BSA . Samples were counter stained with 1 μg/ml DAPI ( Invitrogen , P21490 ) and then washed in TBS supplemented with 3% BSA [59] . Slides were mounted with Vectashield ( Vector Labs , H-1000 ) and visualized using Olympus BX60 Fluorescent Microscope with Spot Insight2 camera at 10x magnification . DNA from fecal contents of wild-type and Stat2-/- infected mice was extracted using the PowerSoil DNA Isolation Kit ( MoBio , 12888–50 ) according to manufacturer’s protocol . High quality isolated DNA was then submitted to SeqMatic for 16S rRNA V4 sequencing using the Illumina MiSeq platform . Data was then analyzed using Qiime pipeline as described [2] . For the co-housing , wild-type ( two or three ) and Stat2-/- mice ( two or three ) were placed in the same cages at the time of weaning and housed together for 5 weeks . The immune cells from the cecal tissue were isolated using the mouse lamina propria dissociation kit ( Miltenyi Biotech , 130-097-410 ) according to manufacturer’s protocol . Briefly , 1x106 cells were resuspended in PBS and stained with live/dead cell discriminator ( Invitrogen , L34597 ) according to manufacturer’s protocol . Cells were then rinsed with PBS , spun down at 400g for 10 minutes and resuspended in 20ml of mouse Fc Block ( Miltenyi , 130-092-575 ) . The cells were then incubated at room temperature for 15 minutes . The mouse Fc Block was left on the cells and the cells were then stained with CD45 PE-Cy7 Rat ( Clone30-F11; Biolegend , 103114 ) , CD3 APC Rat ( Clone 17A2; Biolegend , 100236 ) , B220 APC rat ( Clone RA3-6B2 , Biolegend , 103212 ) , NK1 . 1 APC mouse ( Clone PK136 , Biolegend , 108710 ) and Ly6G Alexa 488 rat ( Clone 1A8; Biolegend , 127626 ) diluted in fluorescence activated cell sorting ( FACS ) buffer according to the manufacturer’s instructions for 30 minutes at 4°C in the dark . The FACS buffer was composed of PBS , 0 . 5% BSA and 2% FBS . Cells were then rinsed with FACS buffer , spun down at 400g for 10 minutes and then resuspended in 100 ml of 4% paraformaldehyde ( BD , 554655 ) , which fixed the cells . Following a 20 min incubation in the dark at room temperature , the cells were washed with FACS buffer . Finally , the cells were spun down and resuspended in 400 ul of FACS buffer . Cells were analyzed on a BD FACS Canto flow cytometer ( BD Biosciences ) and analyzed using FlowJo software ( TreeStar , Inc . , Ashland , OR ) . Mouse bone marrow neutrophils were isolated according to the method of Mocsai et al . [60] . Wild-type and Stat2-/- mice were euthanized and the femur and tibias from the hind legs harvested . Neutrophils were isolated by Percoll density gradient sedimentation , followed by hypotonic lysis to remove erythrocytes . Superoxide anion ( O2- ) generation was measured spectrophotometrically as superoxide-dismutase ( SOD ) -inhibitable cytochrome c reduction . In 96 well plates , bone marrow neutrophils ( 1 . 5 X 106 ) from wild-type and Stat2-/- mice were activated with fMLP ( 10-8M ) in the presence of 5 μg/ml cytochalasin B . For experiments examining the effect of type I IFN on fMLP-stimulated O2- generation , the neutrophils were pre-treated with IFNα ( 1000 units/ml ) prior to the addition of fMLP . The generation of O2- was monitored over a 10 min time-period [61 , 62] . For analysis of bacterial numbers , competitive indices , relative abundance of bacterial populations and fold changes in mRNA levels , values were converted logarithmically to calculate geometric means . Parametric test ( Student t test ) or one-way ANOVA test was used to calculate whether differences were statistically significant ( P < 0 . 05 ) using GraphPad Prism software . All animal experiments were performed in BSL2 facilities with protocols that are approved by AALAC-accredited Temple University Lewis Katz School of Medicine Institutional Animal Care and Use Committee ( IACUC# 4561 ) in accordance with guidelines set forth by the USDA and PHS Policy on Humane Care and Use of Laboratory Animals under the guidance of the Office of Laboratory Animal Welfare ( OLAW ) . The institution has an Animal Welfare Assurance on file with the NIH Office for the Protection of Research Risks ( OPRR ) , Number A3594-01 . | The spread of invading microbes is frequently contained by an inflammatory response . Yet , some pathogenic microbes have evolved to utilize inflammation for niche generation and to support their metabolism . Here , we demonstrate that S . Typhimurium exploits type I IFN signaling , a prototypical anti-viral response , to foster its own growth in the inflamed gut . In the absence of STAT2-dependent type I IFN , production of neutrophil reactive oxygen species was impaired , and epithelial metabolism returned to homeostatic hypoxia . Consequently , S . Typhimurium was unable to respire in the absence of type I IFN , and failed to expand in the gut lumen . Furthermore , perturbation of the gut microbiota was dependent on type I IFN signaling . Taken together , our work suggests that S . Typhimurium utilizes STAT2-dependent type I IFN signaling to generate a niche in the inflamed intestinal tract and outcompete the gut microbiota . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
]
| [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"microbiome",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"salmonellosis",
"bacterial",
"diseases",
"signs",
"and",
"symptoms",
"enterobacteriaceae",
"bacteria",
"neutrophils",
"bacterial",
"pathogens",
"microbial",
"genomics",
"salmonella",
"typhimurium",
"digestive",
"system",
"infectious",
"diseases",
"white",
"blood",
"cells",
"inflammation",
"animal",
"cells",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"salmonella",
"immune",
"response",
"gastrointestinal",
"tract",
"biochemistry",
"diagnostic",
"medicine",
"cell",
"biology",
"anatomy",
"interferons",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"genomics",
"organisms",
"cecum"
]
| 2019 | STAT2 dependent Type I Interferon response promotes dysbiosis and luminal expansion of the enteric pathogen Salmonella Typhimurium |
Temperate phages infect bacteria by injecting their DNA into bacterial cells , where it becomes incorporated into the host genome as a prophage . In the genome of Bacillus subtilis 168 , an active prophage , SPβ , is inserted into a polysaccharide synthesis gene , spsM . Here , we show that a rearrangement occurs during sporulation to reconstitute a functional composite spsM gene by precise excision of SPβ from the chromosome . SPβ excision requires a putative site-specific recombinase , SprA , and an accessory protein , SprB . A minimized SPβ , where all the SPβ genes were deleted , except sprA and sprB , retained the SPβ excision activity during sporulation , demonstrating that sprA and sprB are necessary and sufficient for the excision . While expression of sprA was observed during vegetative growth , sprB was induced during sporulation and upon mitomycin C treatment , which triggers the phage lytic cycle . We also demonstrated that overexpression of sprB ( but not of sprA ) resulted in SPβ prophage excision without triggering the lytic cycle . These results suggest that sprB is the factor that controls the timing of phage excision . Furthermore , we provide evidence that spsM is essential for the addition of polysaccharides to the spore envelope . The presence of polysaccharides on the spore surface renders the spore hydrophilic in water . This property may be beneficial in allowing spores to disperse in natural environments via water flow . A similar rearrangement occurs in Bacillus amyloliquefaciens FZB42 , where a SPβ-like element is excised during sporulation to reconstitute a polysaccharide synthesis gene , suggesting that this type of gene rearrangement is common in spore-forming bacteria because it can be spread by phage infection .
Genetic information is normally preserved across generations in living organisms . However , genomic integrity is sometimes dramatically challenged by DNA rearrangement events , such as homologous recombination , viral genome integration , and transposon spreading . These DNA rearrangements contribute to genetic diversification in the evolutionary history of life on Earth . Importantly , some of these rearrangements are programmed to occur at specific sites and times during cellular differentiation and play crucial developmental roles in a variety of organisms . The best-known example is the rearrangement of immunoglobulin genes in the B lymphocytes of the vertebrate immune system . The assembly in different combinations of the variable ( V ) , diversity ( D ) , and joining ( J ) exons of the immunoglobulin gene generates antigen receptors with extremely diverse binding specificities [1] . DNA rearrangements also modulate gene expression in bacteria during cellular differentiation . For example , during differentiation to a heterocyst , which is a cell type that fixes atmospheric nitrogen , bacteria of the Anabaena genus have the ability to reconstitute the disrupted nifD , fdxN , and hupL genes that are normally inactive in photosynthetic cells [2]–[5] . In the sporulating Gram-positive bacterium Bacillus subtilis , the sigK gene , which encodes the sporulation sigma factor σK , is interrupted by the phage-like element skin . During sporulation , skin is excised and a functional composite sigK gene is produced [6] . B . subtilis cells produce endospores in response to nutrient starvation . The B . subtilis spore envelope is characterized by a succession of concentric layers of chemically distinct composition: the cortex is a peptidoglycan layer assembled between the inner and outer spore membranes , while the coat is an external proteinaceous layer , which can be further subdivided into an inner coat layer and an outer coat layer [7] , [8] . An additional layer called the spore crust was recently discovered outside the outer coat [8]–[10] . Spore formation in B . subtilis has been studied extensively as a model system for cellular differentiation . The process begins with an asymmetric division of the sporulating cell , thus producing two compartments of unequal size , each containing a copy of the genome . The two compartments will differentiate into specific cell types: the forespore and the larger mother cell . During sporulation , a cascade of sporulation sigma factors governs gene expression in a temporally controlled , cell-specific manner [11]–[13] . During the early stages of sporulation , gene expression is controlled by σF in the forespore and σE in the mother cell , whereas σG ( in the forespore ) and σK ( in the mother cell ) control the later stages of the developmental program . The σK-encoding gene , sigK , is disrupted by skin thereby splitting the gene into two protein coding sequences , spoIVCB ( 5′-end of sigK ) and spoIIIC ( 3′-end of sigK ) [6] . A site-specific DNA recombinase , SpoIVCA , promotes excision of skin from the chromosome and the joining in frame of spoIVCB and spoIIIC to reconstitute a functional sigK gene [6] , [14]–[16] . The spoIVCA gene is located in the skin element and is expressed exclusively in the mother cell during sporulation under the control of σE . The rearranged mother cell chromosome is not transmitted to the progeny because the mother cell undergoes autolysis at the end of sporulation to release the mature spore ( whose genome has not been rearranged ) in the environment . A similar rearrangement of the sigK gene was observed in the pathogenic spore-forming bacterium Clostridium difficile [17] . This type of DNA rearrangement was thought to be a unique case because no examples other than sigK had been reported in spore-forming bacteria . However , we recently characterized two other cases of novel intervening sequence elements in mother cell-expressed sporulation genes , vfbin in the spoVFB gene of Bacillus weihenstephanensis KBAB4 and vrin in the spoVR gene of Geobacillus thermoglucosidasius C56-YS93 [18] . These findings suggest that DNA rearrangements may be common in the mother cell genome of spore-forming species , prompting us to embark in a systematic analysis of intervening sequence elements in spore-forming bacteria . B . subtilis 168 contains 10 prophage-like elements [19] . Of these 10 elements , only skin and SPβ are inserted into protein-coding regions . SPβ is integrated into spsM ( spore polysaccharide synthesis protein M ) , thus producing two gene fragments , yodU ( 5′-end of spsM ) and ypqP ( 3′-end of spsM ) . The yodU and ypqP genes are expressed during sporulation under the control of σK [20]–[22] . The most significant difference between skin and SPβ is that skin is a cryptic phage , whereas SPβ is an active prophage . SPβ usually stays in the dormant state ( lysogenic cycle ) . However , when the SOS response is induced by DNA damage , specific genes in the SPβ genome are activated to generate virions that are released after lysis of the host cell ( lytic cycle ) [23] . A putative site-specific recombinase SprA ( SPβ site-specific recombination factor A; formerly yokA ) encoded in the SPβ prophage region is a candidate to promote SPβ excision from the host genome [23] . Nevertheless , the requirement for SprA in SPβ excision had not been investigated until now and the mechanism of excision is poorly understood . In the present study , we examined the fate of SPβ during sporulation . We showed that SPβ was excised from the mother cell genome , thus producing a composite spsM gene . We also investigated the biological function of SpsM and discovered significant changes in the surface properties of spores produced by mutant strains unable to reconstitute a functional spsM gene .
In the genome of B . subtilis 168 , the SPβ prophage is located between two open reading frames ( ORFs ) , yodU ( NCBI gene locus tag BSU19810 ) and ypqP ( BSU21670 ) . Amino acid ( aa ) sequence alignment and comparison to non-SPβ lysogenic B . subtilis strains , such as BEST195 ( NCBI reference sequence no . NC_017196 ) , showed that YodU ( 140 aa ) and YpqP ( 207 aa ) corresponded to the N- and the C-terminal portions of SpsM ( Figure S1A ) . An overlapping 5-aa sequence “TDKAV” was observed at the C-terminus of YodU and at the N-terminus of YpqP . This sequence corresponds to the translation of the nucleotide sequence of the attachment site for SPβ . When the aa sequences of YodU and YpqP were joined at the overlapping sequence , the composite SpsM aa sequence was identical to that of strain BEST195 , thereby indicating that B . subtilis 168 spsM does not contain any mutations ( non-sense , missense , deletions or insertions ) . SpsM is a 341-aa protein , which contains a Polysacc_synt_2 domain ( Pfam accession number , PF02719 ) in the 18–296-aa region . This domain was first observed in Staphylococcus aureus CapD [24] , and is shared among bacterial polysaccharide biosynthesis proteins , such as Campylobacter jejuni WlaL ( putative sugar epimerase/dehydrogenase ) [25] and several sugar epimerases . SpsM shared 38% identity with a B . subtilis paralog , EpsC [26] . EpsC is an UDP–sugar epimerase encoded by the epsC locus and is essential for the production of extracellular polysaccharide ( EPS ) during biofilm formation [26] . B . subtilis SpsM has not been previously characterized , but the conserved domain and similarity to EpsC suggest that SpsM is a sugar epimerase likely to be involved in polysaccharide synthesis . However , a capsular polysaccharide has yet to be identified in vegetative cells of B . subtilis . Considering that yodU ( the 5′-segment of spsM ) and ypqP ( the 3′-segment of spsM ) were identified as sporulation genes in recent transcriptomic analyses of B . subtilis 168 [21] , [22] , we postulated that spsM is involved in the synthesis of the spore polysaccharide . Similar transcriptional profiling results were obtained in the PY79 strain of B . subtilis , which is derived from 168 , but cured of SPβ , where intact spsM was reported as a σK-dependent gene [20] . As a whole , this information led us to hypothesize that in B . subtilis 168 the spsM rearrangement occurs during sporulation to allow production of spore polysaccharide . Figure 1A shows a diagram of the 134-kb long SPβ prophage from the B . subtilis 168 genome . sprA ( formerly yokA; NCBI gene locus tag , BSU21660 ) , which is located immediately upstream of ypqP , encodes a putative site-specific DNA recombinase , which shares 26% identity with SpoIVCA of the skin element . The attachment sites are indicated by triangles . When B . subtilis 168 vegetative cells are treated with mitomycin C ( MMC ) , SPβ is excised ( Figure 1B , left panel ) . Specifically , a wild-type culture was grown in Luria-Bertani ( LB ) medium and MMC ( 0 . 5 µg/ml ) was added to the medium during the early exponential phase of growth [optical density at 600 nm ( OD600 ) = 0 . 25] . DNA samples were extracted from the cells at different time points after MMC addition and digested with NdeI . From 0 to 120 min after MMC treatment , Southern blotting using the sprA-specific probe ( sprA probe ) detected a 9 . 9-kb band ( corresponding to the DNA arrangement before SPβ excision ) . In addition to the 9 . 9-kb band , a second 5 . 6-kb band was detected at 60 , 90 , and 120 min after MMC treatment , which indicated SPβ excision and reconstitution of spsM . Subsequently , to examine spsM rearrangement during sporulation , we performed Southern blotting using DNA samples from sporulating B . subtilis 168 cells ( Figure 1B , right panels ) . The wild-type cells were cultured at 37°C in liquid Difco sporulation medium ( DSM ) and harvested at successive time points one hour before , at the onset of stationary phase and every hour until 8 hours after the onset of stationary phase . Southern blotting using the sprA probe detected the 9 . 9-kb band from T−1 to T8 ( Figure 1B , right top panel ) . The 5 . 6-kb band was detected at T3 and later , thereby indicating that SPβ was excised during sporulation without the need for MMC treatment ( Figure 1B , right top panel ) . We also examined the spsM rearrangement using the ypqP-specific probe ( ypqP probe ) ( Figure 1B , right bottom panel ) . In addition to the 9 . 9-kb band , a 6 . 1-kb band , which corresponded to the composite spsM , was detected at T3 and later . To confirm spsM reconstitution , we determined the DNA sequences at the junction sites of the excised SPβ and composite spsM . The sequencing data showed that SPβ excision in the sporulating cells occurred at the same site as that in the MMC-treated vegetative cells ( Figure S1B ) [23] . The SPβ attachment sites contain 16-bp core sequences ( Figure S1B , nucleotides boxed in red ) and 16-bp inverted repeat sequences ( Figure S1B , arrows ) . Next , we determined the compartment where the spsM rearrangement occurred , i . e . , the mother cell or forespore . The mother cell DNA and forespore DNA were isolated from wild-type cells at T8 and subjected to Southern blotting . The 5 . 6-kb and 6 . 1-kb bands were detected only in the mother-cell compartment , which indicated that SPβ excision during sporulation was a mother cell-specific event and that the SPβ prophage DNA is maintained in the spore genome ( Figure 1C ) . To evaluate the ability of the excised SPβ to form phage particles during sporulation , the supernatant of the DSM culture was filtered and spotted onto a lawn produced by a SPβ-sensitive strain CU1050 [27] . Plaques were not formed ( Figure 1D ) , suggesting that SPβ excised during sporulation is not a phage particle . Nevertheless , we confirmed that the spsM rearrangement can occur during sporulation in a new SPβ lysogen , CU1050 ( SPβ ) , which was obtained by infecting CU1050 cells with the SPβ phage lysate ( Figure 1E ) . This result indicates that the spsM rearrangement system can be transferred to a new host via SPβ phage infection . In addition to B . subtilis 168 , several B . amyloliquefaciens strains carry a prophage sequence similar to SPβ at the spsM locus ( Figure 2A and Table S1 ) . The numbers of SPβ-related genes varied considerably among all of these strains . These SPβ-like elements are likely to be remnants of the SPβ prophage and have probably lost their ability to form infectious phage particles , because large parts of the SPβ-related genes were missing . Since the gene encoding the putative site-specific recombinase , sprA , was conserved in all of these elements , we examined whether the SPβ-like element was excised from the chromosome in B . amyloliquefaciens strain FZB42 ( BGSC catalogue number , 10A6 ) . Figure 2B shows a diagram of the SPβ-like element in B . amyloliquefaciens FZB42 . First , we tested whether the element responded to MMC by analyzing a DNA sample prepared from MMC-treated vegetative cells of strain FZB42 and subjected to Southern blotting . The B . amyloliquefaciens sprA-specific probe ( sprABam probe ) detected a single 5 . 9-kb band from 0 to 120 min after MMC addition , indicating that excision of the element did not occur in the MMC-treated vegetative cells ( Figure 2C , upper panel ) . Subsequently , Southern blotting was performed using a DNA sample obtained from sporulating cells of strain FZB42 ( Figure 2C , lower panels ) . Bands indicating excision of the element ( 13 kb , left panel ) and the generation of the composite spsM ( 3 . 6 kb , right panel ) were detected using the sprABam and ypqPBam probes , respectively . These data indicate that the SPβ-like element of B . amyloliquefaciens FZB42 exhibits a behavior distinct from the B . subtilis SPβ , but similar to skin , because it did not respond to MMC treatment and was excised only during sporulation . Considering that in B . subtilis SPβ excision occurs both during sporulation and in response to DNA damage , whereas in B . amyloliquefaciens excision of the SPβ-like element only occurs during sporulation , it is likely that different mechanisms control prophage excision during sporulation and upon MMC treatment . To analyze how SPβ controlled its excision and to determine whether spsM expression always followed prophage excision , we constructed transcriptional lacZ fusions to yodU ( 5′-spsM ) [YODUd; yodU::pMutinT3 , PyodU–lacZ] and sprA ( SPRAd; sprA::pMutinT3 , PsprA–lacZ ) using the pMutinT3 insertion plasmid ( Figure S2A ) . Insertion of the pMutinT3 vector into a genome locus causes inactivation of the corresponding gene and allows analysis of its expression profile by measuring β-galactosidase activity , because the gene of interest is now transcriptionally fused to lacZ [28] . In the YODUd strain , PyodU–lacZ was expressed during the late stages of sporulation , consistent with the previously reported σK-dependency for yodU expression ( Figure 3A , left panel , hour 8 and later ) . The timing of expression of yodU was delayed by 2 hours when compared to that of cotG , another σK-dependent gene [29] . This delay is likely due to the fact that yodU expression also requires the transcription factor GerE , which regulates gene expression in the mother cell during the ultimate stage of sporulation , as previously shown [20] . By contrast , PyodU–lacZ was not expressed in MMC-treated vegetative cells ( Figure 3B , left panel ) , indicating that prophage excision does not systematically trigger spsM expression . Analysis of the SPRAd mutant strain by Southern blotting did not reveal any difference in the band patterns of vegetative and sporulating cells ( Figures 4A , middle panels and S2A ) , showing that sprA was necessary for spsM reconstitution . Nicolas et al . predicted a putative binding site for the housekeeping σ factor σA at positions −85 to −57 ( TTGTTT for the −35 box and TAAAAT for the −10 box ) relative to the sprA start codon [22] . Consistent with a σA-dependent pattern of expression , but at odds with a specific role for SprA during the late stages of sporulation , the PsprA–lacZ activity kept increasing during vegetative growth , peaked during the early stages of sporulation and gradually decreased as sporulation proceeded ( Figure 3A , middle panel ) . The sprA expression level in vegetative cells was not increased by MMC addition ( Figure 3B , middle panel ) . These unexpected results suggest that an additional factor ( s ) regulates the timing of prophage excision during sporulation and following DNA damage . We observed that sprB ( formerly yotN; NCBI gene locus tag , BSU19820 ) , a SPβ gene located downstream of yodU , was conserved in all of the SPβ-like elements ( Figure 2A ) . It encodes a 58-aa protein with no significant similarity to characterized proteins . To test whether sprB was required for excision , we constructed a sprB deletion mutant strain ( SPRBd ) . Southern blotting revealed that SPRBd was defective in SPβ excision ( Figures 4A , right panels and S2B ) , indicating that sprB was necessary for excision . As expected , PsprB–lacZ was expressed during the middle and late stages of sporulation ( Figure 3A , right panel ) and was also induced by MMC addition to vegetative cells ( Figure 3B , right panel , 60 min and later ) . To examine the correlation between SPβ excision and sprA and sprB expression , we constructed the sprA-inducible strain ( BsINDA ) and the sprB-inducible strain ( BsINDB ) , where sprA or sprB expression can be induced by isopropyl β-D-1-thiogalactopyranoside ( IPTG ) addition . SPβ was excised when sprB expression was induced in BsINDB , but not when sprA was over-expressed in BsINDA ( Figure S3 ) . Combined with the results from Figure 3 , showing that PsprA–lacZ is expressed at significant levels during vegetative growth and the early stages of sporulation , we conclude that expression of sprA alone is not sufficient to excise SPβ from the chromosome ( Figure S3A ) . By contrast , when sprB is induced , either in the presence of MMC , during sporulation , or artificially by IPTG addition , excision of SPβ will ensue ( Figure S3B ) , provided that SprA is also present . In summary , both sprA and sprB are necessary for excision , but the temporal control of excision is dependent on sprB . To determine whether SPβ genes other than sprA and sprB were also required for excision , we constructed a SPβ mutant strain ( SPmini ) , where all the SPβ genes were deleted , except sprA and sprB . The SPmini strain retained the capacity for spsM rearrangement during sporulation ( Figure 4C , left panels ) , indicating that sprA and sprB are necessary and sufficient for SPβ excision during sporulation . By contrast , SPmini did not undergo excision upon MMC treatment ( Figure 4C , right panels ) , suggesting that an additional gene ( s ) or regulatory sequence present in SPβ but absent in SPmini may be required to promote excision and/or trigger sprB expression following DNA damage . Since sprB is a key factor in the control of SPβ excision , we analyzed its transcriptional regulation ( Figure 5 ) . We performed Northern blotting using a sprB-specific probe ( Figure 5A , thick black line ) . A major band of 5 . 0 kb and minor bands of 1 . 2 and 2 . 0 kb were detected in MMC-treated vegetative cells but not in untreated cells ( Figure 5C , columns 1 and 2 ) . By contrast , a single 0 . 2-kb band was detected during sporulation ( Figure 5C , column 3 ) . This result suggested that sprB was transcribed from distinct promoters upon MMC treatment and during sporulation . Lazarevic et al . reported that the yosX gene , which is located 5 kb upstream of sprB , possesses a σA-dependent promoter [23] , but no other σA-dependent promoter was predicted between yosX and sprB ( Figure 5A ) . Thus , it is likely that the major 5 . 0 kb band detected upon MMC treatment corresponds to a transcript originating from the yosX promoter , while the minor bands could correspond to truncated transcripts . Next , we performed RT-PCR using a sprB-specific reverse transcription primer ( Figure 5A , RT primer , red arrow ) followed by PCR amplification of the sprB cDNA using yosX , yotBCD , or sprB specific primers ( Figure 5A , black arrows ) . When the sprB cDNA was obtained from the MMC-treated cells ( Figure 5D , column 2 ) the yosX , yotBCD , and sprB regions were successfully amplified , whereas when the sprB cDNA was obtained from sporulating cells , only the sprB region could be amplified ( Figure 5D , column 3 ) . This result indicates that sprB is indeed co-transcribed with the upstream genes upon MMC treatment while it appears to be monocistronically transcribed during sporulation . To determine the 5′ end of the sprB transcript during sporulation , we carried out 5′ RACE PCR with total RNA extracted at T4 . The sprB transcriptional start site ( TSS ) was found to be located 20 nt upstream of the start codon ( Figures 5B and S4 ) . Using DBTBS Search Tools ( http://dbtbs . hgc . jp/ ) [30] , a putative σE- or σK-binding site was found directly upstream of the TSS of sprB ( Figure 5B ) . However , while the −10 element was a perfect match to the σE- or σK-consensus sequence , the putative −35 element of the sprB promoter was an imperfect match ( Figure 5B ) . It is therefore possible that an additional mother cell transcription factor , such as SpoIIID , GerR or GerE , is required along with σE or σK for optimal expression of sprB . To test whether sprB expression is restricted to the mother cell , as would be expected if it is controlled by a σE or σK , we constructed strain BsSPRBG , which harbors a plasmid carrying the translational fusion sprB–gfp without the upstream phage genes . As expected , GFP fluorescence in BsSPRBG was detected only in the mother cell ( Figure 5E ) . Importantly , this observation is also consistent with the data presented above ( Figure 3A ) , where PsprB–lacZ activity was detected during the middle to the late stages of the sporulation , when σE and σK are most active . To investigate the functional role of spsM in sporulation , we used the YODUd ( yodU ) and SPRAd ( sprA ) strains . Since SPRBd exhibited the same phenotype as SPRAd , only the SPRAd strain will be considered further . We analyzed the morphologies of wild-type , YODUd , and SPRAd spores using phase-contrast microscopy and a negative staining procedure . When the spores were negatively-stained with Indian ink [31] , which is a stain commonly used to reveal polysaccharide capsules , a clear halo was visible around the wild-type spores , but not around the YODUd and SPRAd spores ( Figure 6A , top panels ) . The appearance of a halo is consistent with the presence of polysaccharides around the wild-type spore . Introduction of the composite spsM gene at the amyE locus of the mutant strains complemented the sprA and yodU mutations ( SPRAc and YODUc ) in the sense that the halo was restored ( Figure 6A , top panels , sprA spsM+ and yodU spsM+ ) . In addition , we observed that this putative polysaccharide layer of the wild-type spore was loose , because it can easily be removed from the spores by boiling in a buffer containing SDS . After this treatment , the wild-type and spsM+ spores became indistinguishable from the SPRAd and YODUd spores , as none of the spores exhibited a halo ( Figure 6A , bottom panels ) . These results suggest that the composite spsM is necessary for the production of an external spore structure most likely composed of polysaccharides . Surface extracts from wild-type spores were loaded on a 5% polyacrylamide gel , separated by electrophoresis and stained with stains-All , a cationic carbocyanine dye that stains polysaccharides , nucleic acids , and acidic proteins . The spore surface component was detected as a bright blue band ( Figure 6B ) , which indicated the presence of a high molecular weight substance . The blue band was not detected in extracts from SPRAd and YODUd ( Figure 6B , sprA and yodU ) , whereas it was detected in extracts from SPRAc ( sprA spsM+ ) and YODUc ( yodU spsM+ ) . These results imply that the formation of the spore surface component is dependent on the function of the composite spsM . SpsM is a paralog of a polysaccharide synthesis protein , EpsC [26] . Thus , the high molecular weight substance from the spore surface is likely to be a polysaccharide , whose synthesis and/or attachment to the spore is dependent on SpsM . In addition , this high molecular weight substance was inferred to be produced in the mother cell because the composite SpsM protein fused to GFP was observed to reside in the mother cell during sporulation ( Figure S5 ) , consistent with its regulation by σK [20] . We quantified the amount of spore surface component using the method described by Hammerschmidt et al . [32] . The levels of high molecular weight substance in the SPRAd and YODUd spore surface extracts decreased to 12 . 5% and 5 . 0% of the amount isolated from wild-type spores ( Figure 6C ) . Next , we analyzed the monosaccharide composition of the wild-type spore surface extract . The extract was hydrolyzed and fluorescently labeled with 4-amino-benzoic acid ethyl ester ( 4-ABEE ) . HPLC analysis detected three major peaks . By comparison to fluorescently labeled monosaccharide standards , we infer that the two peaks detected in the extracts at retention times of 10 . 9 and 30 . 8 min corresponded to galactose and rhamnose , respectively ( Figure S6 , peaks 3 and 12 ) . A peak at 6 . 4 min , which did not correspond to any monosaccharide standard , was considered to be an unknown monosaccharide ( s ) or could result from an incomplete hydrolysis of oligosaccharides . The galactose and rhamnose peaks accounted for 21 . 1% and 68 . 1% of total amount of monosaccharides detected by HPLC , respectively . The presence of rhamnose at the spore surface has been previously reported and was shown to be dependent on the enzymes SpsI , SpsJ , SpsK and SpsL , whose synthesis is dependent on σK during sporulation [20] , [33] , [34] . In conclusion , our experiments indicate that polysaccharides are present in spore surface extracts and that spsM is involved in their production and/or attachment to the spore envelope . Subsequently , we investigated the functional roles of the spore polysaccharides . YODUd and SPRAd retained the ability to produce phase-bright and wet-heat resistant spores although spore titers in DSM cultures were slightly smaller than that of the wild type ( Table S2 ) . In addition , a SPβ-cured strain , SPless , produced normal wet-heat resistant spores with a sporulation efficiency that was comparable to that of the wild type ( Table S2 ) . However , we noticed that the mutant spores exhibited significant differences in their properties . The purified mutant spores formed aggregates and displayed enhanced adhesion to solid surfaces , such as borosilicate glass and polypropylene . Figure 7A reports that mutant spores adhere to Pyrex tubes ( 13×100 mm , Corning ) , whereas wild-type and spsM+ spores do not . Figure 7B shows the result of an adhesion test using polypropylene tubes ( see Materials and Methods ) . While 80%–90% of YODUd and SPRAd spores had adhered to the tubes after five transfers , wild-type and spsM+ spores barely adhered to the tube even after ten transfers . Finally , we investigated the adhesive properties of the mutant spores on DSM-agar plates ( Figure 7C ) . B . subtilis cells were cultured at 37°C on DSM plates for a week to allow sporulation . After this period , >95% cells on the plates became mature spores ( Figure 7C , upper panels ) . After the plate was rinsed with water , the wild-type spores dispersed in water and disappeared from the plate ( Figure 7C , lower panels ) . However , the SPRAd and YODUd spores were barely resuspended in water , and most of the spores were left on the plates . Therefore , our results suggest that the spore polysaccharide are beneficial for the dispersal of B . subtilis spores through water and help prevent adhesion to certain types of surfaces .
We demonstrated that both B . subtilis and B . amyloliquefaciens reconstitute a functional spsM gene during sporulation through developmentally-controlled excision of the SPβ prophage ( Figures 1 and 2 ) ; however , while SPβ is an active prophage in B . subtilis , it has become a cryptic prophage in strains of B . amyloliquefaciens ( Figure 2 and Table S1 ) . The observation that the spsM rearrangement system can be transferred to a non-lysogenic strain via SPβ infection ( Figure 1E ) suggests that the element was originally acquired by the current lysogenic strains following an infection with an ancestral phage identical or very closely related to SPβ . We speculate that the strains of B . amyloliquefaciens have been infected with SPβ earlier than B . subtilis and have since lost most of the original phage genes , probably because they did not confer significant advantages or may even be harmful to the host ( Figure 8A ) . Prophage-mediated reconstitution of mother cell-specific sporulation genes is a common event since other intervening elements ( e . g . , skin , vfbin , and vrin ) that carry phage-related genes have been previously observed in several spore-forming bacteria [18] , [35] . Similar to B . amyloliquefaciens SPβ , these elements are the descendants of ancestral prophages and have now become defective for producing phage particles , but are still being excised under specific conditions ( Figure 8A ) . Importantly , the excision of these elements from the host genome is developmentally regulated and confined to a terminally differentiated cell type , the mother cell ( Figure 1C ) [6] , [16] , [18] . Limiting the DNA rearrangement to the mother cell genome ensures that the phage DNA is maintained in the spore genome ( Figure 8B ) . Thus , after spore germination , SPβ is vertically transferred to the progeny upon cell division as a permanent element in the host genome sequence . Insertion of prophages in sporulation genes is advantageous to the host for at least two reasons: ( 1 ) to add one level of control to the progression of sporulation; and ( 2 ) to acquire immunity against other phages . As an example of the first type of benefit , the presence of skin in the host genome was shown to be required for efficient sporulation in Clostridium difficile [17] , even though it is dispensable in B . subtilis [36] . During sporulation in B . subtilis , the temporal control of σK activity is achieved by triggering the proteolytic removal of an inhibitory pro-sequence at its N-terminus [37] . Since σK does not possess the pro-sequence in C . difficile , another regulatory mechanism is required to control the timing of σK activation [17] . Regarding the second type of benefit , phages constitute an ideal vehicle for the host to acquire genes that provide selective advantages , especially as protection against other phage infections . In addition , sporulation genes are suitable locations for bacterial attachment sites because they are not essential for vegetative cell growth and viability . In general , lysogenic bacteria become immune to further infections by acquiring the ability to synthesize repressor proteins for closely related phages . Furthermore , in the case of SPβ , the prophage carries both sunA , which encodes sublancin , an antimicrobial that inhibits cell growth of non-SPβ lysogens [38] , and nonA , which confers resistance to infections by the virulent phage SP10 [39]–[41] . Since the SPβ-cured strain , SPless , produces normal spores ( Table S2 ) , the presence of SPβ in the B . subtilis genome is more likely to be beneficial to the host by providing immunity against other phages rather than adding a layer of control to sporulation progression . Recently , Rabinovich et al . have reported a similar prophage excision event in Listeria monocytogenes [42] . In this case , a functional comK gene is reconstituted to favor escape from phagocytosis . This observation suggests that prophage-mediated gene reconstitution is common among bacteria and is not limited to spore-formers . Of the SPβ genes , we found that only sprA and sprB were required for excision ( Figure 4 ) . We have shown that sprB was expressed in response to MMC treatment ( DNA damage ) and is developmentally regulated during sporulation , whereas sprA was expressed irrespective of the host cell status ( Figure 3 ) . Moreover , overexpression of sprB , but not of sprA , successfully promoted spsM reconstitution in vegetative cells , even without MMC induction ( Figure S3 ) . Our results suggest that sprB is the factor that controls the timing of SPβ excision . SprA belongs to a family of large serine recombinases , which rely on recombination directionality factors ( RDFs ) to promote excision [43] . RDFs are small DNA-binding proteins that initiate the assembly of the recombinase–DNA complexes . SprB may serve as a RDF for the SprA recombinase during SPβ excision . Lazarevic et al . found putative SPβ repressor-binding sites called SPBRE in the promoter regions of yorE , yorM , yorZ , and yosX [23] . Repression is expected to be relieved upon MMC treatment . Thus , activation of sprB expression in response to DNA damage seems to result from derepression of yosX and its downstream genes , which include sprB . Importantly , SPmini is not subjected to excision upon MMC treatment ( Figure 4C ) , since it is lacking the SOS-inducible phage genes upstream of sprB . In addition , we identified a mother cell-specific promoter immediately upstream of the sprB gene ( Figures 5AB ) . We propose that the reason why phage particles are not produced after SPβ excision during sporulation is because many SPβ genes lack sporulation-specific promoters , resulting in insufficient production of phage structural components . After SPβ excision , transcription of spsM is controlled by σK [20] . The σK-encoding gene , sigK , is itself generated by excision of skin [6] . Therefore , expression of the composite spsM requires two DNA rearrangement events mediated by the phage elements SPβ and skin . Our study also revealed an important connection between spsM function and B . subtilis spore surface properties . In Bacillus anthracis and Bacillus cereus strains , spores are surrounded by an exosporium , which is a loose-fitting and balloon-like structure , containing glycoproteins and polysaccharides [44] , [45] . The exosporium is not observed in B . subtilis spores , but the crust can be considered to be an exosporium-like structure , even though it does not display the balloon-like structure of a typical exosporium . The protein composition of the crust has been characterized to some extent , in the sense that the coat proteins CgeA , CotG , and CotXYZ were identified as crust components [9] , [10]; however , beyond the fact that rhamnose , whose synthesis is dependent on spsIJKL [34] , is a component of the spore surface [33] , the spore polysaccharide composition in B . subtilis remains poorly characterized . Our analyses indicate that the B . subtilis spore polysaccharide also comprises galactose in addition to rhamnose and possibly another monosaccharide of unknown identity . In addition , we have shown that the production and/or attachment of the polysaccharide to the spore surface were spsM-dependent ( Figure 6 ) . Our results also indicated that although spsM mutant spores were as heat-resistant as wild type spores ( Table S2 ) , they were considerably more sticky and aggregated in water ( Figure 7 ) . The slight reduction of the spore numbers in the spsM mutant strains ( Table S2 ) may be due to their increased adhesive properties . The hydrophobic phenotype of the spsM mutant spores may be attributable to the hydrophobic nature of the crust proteins CotXYZ and CgeA proteins [46] . In the absence of polysaccharide addition , these proteins become directly exposed at the spore surface , and the consequence may be a decrease in the solubility of spores in water . In natural environments , water flow , such as rainfall , rivers , and sea currents , is likely to play a role in spore dispersal . For an immotile spore , the ability to be transported to a different niche , where it can germinate and resume growth , constitutes a major advantage . In conclusion , B . subtilis SPβ prophage has two pathways to excision . In response to host DNA damage , the SPβ prophage is excised from the host genome to form phage particles . By contrast , during sporulation , SPβ excision occurs in the mother cell to reconstitute a sporulation gene , spsM , a necessary event for spore polysaccharide synthesis . Although phage particle formation does not occur during sporulation , the SPβ prophage is propagated vertically to the progeny because phage excision is limited to the mother cell genome .
The primers used in this study are shown in Table S3 . The bacterial strains and plasmids used in this study are listed in Table S4 . Standard genetic manipulations of B . subtilis were performed as previously described [47] . Internal segments of yodU ( +28 to +244 relative to the first nucleotide of the start codon ) , sprA ( +29 to +990 ) , and cotG ( +21 to +217 ) were amplified from the chromosome of B . subtilis 168 using primer pairs P01/P02 , P03/P04 , and P05/P06 , respectively . PCR products were digested with HindIII and BamHI , and inserted into the HindIII–BamHI site of pMutinT3 . The resulting pMUT-yodU , pMUT-sprA , and pMUT-cotG , plasmids were introduced into B . subtilis 168-competent cells to disrupt yodU , sprA , and cotG , respectively . The resulting YODUd , SPRAd , and COTGd strains were selected on Luria-Bertani ( LB ) agar plates containing 0 . 3 µg/ml erythromycin . To construct BsINDA and BsINDB , the 5′ portions containing the SD sequence of sprA ( −27 to +990 ) and of sprB ( −20 to +89 ) were amplified using primer pairs P07/P04 and P08/P09 , respectively . PCR products were digested with HindIII and BamHI and inserted into the HindIII–BamHI site of pMutinT3 , which allows generation of a fusion transcript with a gene encoding β-galactosidase and placing genes downstream of an IPTG-inducible promoter ( Pspac ) . The resulting plasmids , pMUT-sprAind and pMUT-sprBind , were introduced into B . subtilis 168-competent cells . The transformants were selected on LB-agar plates containing 0 . 3 µg/ml erythromycin . To obtain the SPβ-cured strain ( SPless ) , we cultivated BsINDB at 37°C in LB liquid medium in the presence of 0 . 5 mM IPTG overnight . The culture was spread on a LB-agar plate after dilution with fresh LB medium . The plate was incubated at 37°C overnight . The next day , SPβ-cured colonies were selected by colony PCR using primer pair P10/P11 and by erythromycin sensitivity . A sprB-deletion mutant ( SPRBd ) and a strain harboring the minimized SPβ ( SPmini ) were constructed by double-crossing over recombination using the ermC gene cassettes . To construct SPRBd , DNA fragments corresponding to the upstream ( −1126 to −1 ) and the downstream ( +169 to +2246 ) flanking regions of sprB were amplified from the B . subtlis 168 genome using primer pairs P12/P13 and P14/P15 . A DNA fragment containing the ermC gene was amplified from a pUCE191 plasmid vector using primer pair P16/P17 . The DNA fragments were combined by over-extension PCR ( OE-PCR ) using the primer set P12/P15 . The resulting PCR product was introduced into B . subtilis-168 competent cells and by double crossing-over replacement of the sprB locus by the ermC cassette . The transformants were selected on the LB-agar plates containing 0 . 3 µg/ml erythromycin . For construction of the SPmini strain , the primer sets P18/P19 and P20/P04 were used for amplification of the DNA fragments containing the sprB gene ( −331 to +1301 relative to the first nucleotide of the sprB start codon ) and the sprA gene ( −84 to +990 relative to the first nucleotide of the sprA start codon ) with their promoter regions and the attachment sites . The DNA fragments were combined with the ermC cassette by OE-PCR with the primer set P19/P04 and used for transformation of B . subtilis 168 . Transformants were selected on erythromycin-containing LB plates . DNA fragments containing a composite spsM with its promoter region ( −374 to +1080 , relative to the first nucleotide of the start codon of the composite spsM ) were amplified from chromosomal DNA of B . subtilis 168 sporulating cells using primer pair P21/P22 . The PCR product was digested with EcoRI and BglII , and inserted into the EcoRI–BamHI site of the integration vector pMF20 [48] . The resulting plasmid pMFspsM was linearized by BglII-digestion and subsequently integrated into amyE locus of YODUd and SPRAd by double crossover recombination . The resulting YODUc and SPRAc strains were selected on LB agar plates containing 0 . 3 µg/ml erythromycin and 5 µg/ml chloramphenicol . Overnight cultures of B . subtilis strains grown at 37°C in liquid LB medium were diluted 1∶100 with fresh liquid Difco sporulation medium ( DSM ) and incubated at 37°C with shaking . The CU1050 derivatives did not sporulate well in liquid DSM . Therefore , these strains were induced to sporulate on DSM-agar plates . One hundred microliter of overnight cultures of the CU1050 derivatives in LB medium were spread on 90-mm DSM-agar plates and incubated at 37°C . B . subtilis and B . amyloliquefaciens strains were cultured at 37°C in liquid DSM . We harvested 4 ml of the culture by centrifugation at various time points during sporulation . For induction of sporulation of the CU1050 derivatives , cells were spread on DSM-agar plates and incubated at 37°C for either 3 ( vegetative phase ) or 12 hrs ( sporulation phase ) . Cell morphology was monitored by phase-contrast microscopy . After addition of 20 ml of deionized distilled water ( DDW ) to the plates , the cells were gently scraped from the plates and harvested by centrifugation . Genomic DNA was extracted as follows: cell pellets were suspended in 500 µl of TEN buffer [10 mM Tris-HCl ( pH 7 . 5 ) , 10 mM EDTA , and 0 . 1 M NaCl] containing 250 µg/ml lysozyme and 10 µg/ml RNase A . The suspension was incubated at 37°C for 20 min , supplemented with 0 . 1% of sodium dodecyl sulfate ( SDS ) , and incubation was continued for 5 min . Genomic DNA was isolated by phenol extraction and precipitated by ethanol . The DNA pellet was resolved in TE buffer [10 mM Tris-HCl ( pH 8 . 0 ) and 1 mM EDTA] . To isolate B . subtilis forespore DNA , 50 ml of the DSM culture at T8 were harvested by centrifugation . The cell pellets were resuspended in TEN buffer containing 250 µg/ml lysozyme and 100 µg/ml DNase I and incubated at 37°C for 20 min to lyse the mother cells and non-sporulating cells . The suspension was centrifuged and the pellet washed five times by resuspension and recentrifugation in 2 ml of TEN buffer . The forespore pellet was resuspended in SUTD buffer [1% ( w/v ) SDS , 8 M Urea , 50 mM Tris-HCl ( pH 8 . 0 ) , and 50 mM dithiothreitol] [47] , [49] and incubated at 37°C for 90 min . The suspension was washed five times in 2 ml of TEN buffer . The forespore pellet was lysed with 250 µg/ml lysozyme , followed by phenol extraction and ethanol precipitation . The spore DNA pellet was resuspended in TE buffer . To prepare the DIG-labeled probes , DNA fragments corresponding to the 358-bp ypqP probe , the 982-bp sprA probe , and the 535-bp sprB probe were amplified from the chromosomal DNA of B . subtilis using the primer pairs P23/P24 , P03/P04 , and P18/P25 , respectively . DNA fragments corresponding to the 600-bp sprABam and the 500-bp ypqPBam were amplified from the chromosomal DNA of B . amyloliquefaciens FZB42 using the primer pairs P26/P27 and P28/P29 , respectively . The resulting PCR products were gel-purified and labeled using DIG-High Prime ( Roche ) according to the supplier's instructions . Chromosomal DNA ( 2 . 5 µg ) was digested with 20 U of restriction enzymes at 37°C for 16 hours , separated by 0 . 8% agarose gel electrophoresis and blotted onto a Hybond-N+ membrane ( GE Healthcare ) using Alkaline solution ( 10× SSC and 0 . 2 N NaOH ) . Hybridization and detection were performed according to the DIG Application Manual ( Roche ) . Signals were detected by a nitro-blue tetrazolium/5-bromo-4-chloro-3-indolyl-phosphate ( NBT/BCIP ) reaction using the DIG Nucleic Acid Detection Kit ( Roche ) . YODUd ( yodU::pMutinT3 , PyodU–lacZ ) , SPRAd ( sprA::pMutinT3 , PsprA–lacZ ) , and BsINDB ( sprB::pMutinT3 , PsprB–lacZ , Pspac–sprB ) were used to monitor the yodU , sprA , and sprB promoter activities , respectively . The B . subtilis strains were sporulated at 37°C in liquid DSM . The samples were collected at various time points after the end of the exponential phase of growth . β-galactosidase activity was determined using the method described by Miller [50] . B . subtilis 168 cells were grown at 37°C in 50 ml LB medium up to the early log phase ( OD600 = 0 . 25 ) . The culture was further incubated at 37°C for 60 min in the presence or absence of MMC ( 0 . 5 µg/ml ) , and harvested by centrifugation . For preparation of the sporulating cells , the B . subtilis 168 cells were cultured at 37°C in 50 ml of liquid DSM and harvested at T4 by centrifugation . Total RNA was isolated as described previously [51] . Five micrograms of total RNA were mixed with two volumes of denaturing buffer [50% formamide , 6% formaldehyde , 20 mM morpholinopropanesulfonic acid ( MOPS ) ( pH 7 . 0 ) , 5 mM sodium acetate , 1 mM EDTA , 0 . 05% bromophenol blue , and 10% glycerol] and incubated at 55°C for 10 min . The denatured RNA sample was loaded to a 2% denaturing agarose gel containing 2% formaldehyde , separated by electrophoresis in 1× MOPS buffer [20 mM MOPS ( pH 7 . 0 ) and 5 mM sodium acetate] , and capillary-transferred to Hybond N+ membrane ( GE Healthcare ) overnight in 10× SSC buffer [1 . 5 M NaCl and 150 mM sodium citrate ( pH 7 . 0 ) ] . The resulting membrane was baked at 80°C for 2 hrs and stained with a methylene blue solution [0 . 03% methylene blue and 0 . 3 M sodium acetate ( pH 5 . 6 ) ] . Hybridization and detection were performed according to the DIG Application Manual ( Roche ) . Signals were detected using CDP-Star ( Roche ) . Total RNA from B . subtilis 168 vegetative cells with or without MMC treatment and sporulation cells at T4 were prepared as described above . The sprB cDNA was synthesized from 5 µg of the total RNA by an AMV reverse transcriptase XL ( Takara ) using the sprB-specific primer P30 , according to the manufacturer's instructions . Internal segments of the sprB , yosX , and yotBCD coding regions were amplified from the cDNA by 25-cycled PCR reactions using ExTaq ( Takara ) and the primer sets P08/P09 , P31/P32 , and P33/P34 , respectively . PCR products were analyzed by 2% agarose gel electrophoresis . To construct the pUBsprBgfp plasmid carrying the sprB gene translationally fused to gfp , a DNA fragment containing sprB ORF and its sporulation-specific promoter was amplified from the B . subtilis 168 chromosomal DNA using primers P18/P25 . An 858-bp DNA fragment of gfp was amplified from the pMF20 vector [48] using primer pair P35/P36 . The pUB110 plasmid vector [52] was linearized by PCR using the primer set P37/P38 . The sprB DNA fragment , the gfp DNA fragment and the linearized pUB plasmid were combined by OE-PCR and amplified using the primer set P18/P37 . The PCR product was self-ligated with T4 DNA ligase ( Takara ) in the presence of polynucleotide kinase ( Takara ) and introduced into B . subtilis 168-competent cells as described previously [18] . The transformants were selected by addition of 10 µg/ml kanamycin on LB-agar plates . B . subtilis strains , BsSPRBG and BsSPSMG , were cultured at 37°C in liquid DSM containing FM4-64 ( 0 . 25 µg/ml ) . For the cultivation of the BsSPRBG strain carrying the pUBsprBgfp plasmid , kanamycin was added to the medium at a final concentration of 10 µg/ml . Sporulating cells were observed using fluorescence microscopy as previously described [53] . Overnight cultures of B . subtilis strains in LB medium were spread on 90-mm DSM-agar plates . The DSM plates were incubated at 37°C for 6 days and kept at 4°C for a day . After the addition of 20 ml of DDW to the plates , the spores were gently scraped from the plates . The spores were centrifuged at 8 , 000× g for 30 min . The spore pellets were resuspended in 20 ml of DDW and kept overnight at room temperature . The spores were further purified as described by Carrera et al . [54] . B . subtilis spores were negatively stained with Indian ink , as previously described [31] with a slight modification . The purified spores were resuspended in DDW , and 2 µl of the suspension was mixed with an equivalent volume of Indian ink ( Daiso Sogyo , Japan ) on a slide glass . A cover glass was placed on the slide glass and any excess fluid was pushed out using thumb pressure . The negatively stained outermost layer of the spore was observed using phase-contrast microscopy . The purified spores were resuspended in DDW and the final OD600 was adjusted to 50 . Next , 100 µl of the spore resuspension were added to 100 µl of the SDS buffer [125 mM Tris–HCl ( pH 6 . 8 ) , 2% SDS , and 4% β-mercaptoethanol] and incubated at 98°C for 10 min . The supernatant was collected after centrifugation at 20 , 400× g for 5 min and 40 µl of the supernatant was loaded onto a 5% native polyacrylamide gel , which was separated by electrophoresis at 100 V for 30 min in 1× TBE buffer [44 . 5 mM Tris , 44 . 5 mM borate , and 1 mM EDTA ( pH 8 . 0 ) ] . The resulting gel was stained overnight at room temperature using dye solution [0 . 025% ( w/v ) Stains-All ( Sigma-Aldrich ) , 7 . 5% formamide , 3% acetic acid , and 25% 2-propanol] . The polysaccharide components in the spore surface extract were quantified as described by Hammerschmidt et al . [32] . The monosaccharide composition of the spore polysaccharide was determined as described in Supporting text S1 . The purified B . subtilis spores were resuspended in DDW , and the final OD600 was adjusted to 0 . 5 . Each 1 ml of the spore resuspensions was added to a polypropylene tube ( 8 . 8×40 mm; Safe-Lock tube 2 . 0 ml; Eppendorf ) . The spore resuspension was vortexed gently for 10 s and transferred to a fresh polypropylene tube . This operation was repeated 10 times . Total amount of the spores bound to the tubes [Adhesion ( % ) ] was calculated from the percentage decrease in OD600 of the spore resuspension as follows: 100×[ODi−ODn]/ODi , where ODi and ODn are the initial OD600 ( = 0 . 5 ) and OD600 of each binding reaction , respectively . | Integration of prophages into protein-coding sequences of the host chromosome generally results in loss of function of the interrupted gene . In the endospore-forming organism Bacillus subtilis strain 168 , the SPβ prophage is inserted into a previously-uncharacterized spore polysaccharide synthesis gene , spsM . In vegetative cells , the lytic cycle is induced in response to DNA damage . In the process , SPβ is excised from the genome to form phage particles . Here , we demonstrate that SPβ excision is also a developmentally-regulated event that occurs systematically during sporulation to reconstitute a functional spsM gene . Following asymmetric division of the sporulating cell , two cellular compartments are generated , the forespore , which will mature into a spore , and the mother cell , which is essential to the process of spore maturation . Because phage excision is limited to the mother cell genome , and does not occur in the forespore genome , SPβ is an integral part of the spore genome . Thus , after the spores germinate , the vegetative cells resume growth and the SPβ prophage is propagated vertically to the progeny along with the rest of the host genome . Our results suggest that the two pathways of SPβ excision support both the phage life cycle and normal sporulation of the host cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"bacteriology",
"gram",
"positive",
"bacteria",
"viruses",
"biochemistry",
"bacterial",
"physiology",
"bacteriophages",
"bacterial",
"genes",
"genetics",
"biology",
"and",
"life",
"sciences",
"dna",
"microbiology",
"dna",
"recombination",
"bacterial",
"spores",
"organisms"
]
| 2014 | Developmentally-Regulated Excision of the SPβ Prophage Reconstitutes a Gene Required for Spore Envelope Maturation in Bacillus subtilis |
Middle East respiratory syndrome coronavirus ( MERS-CoV ) causes severe respiratory infections that can be life-threatening . To establish an infection and spread , MERS-CoV , like most other viruses , must navigate through an intricate network of antiviral host responses . Besides the well-known type I interferon ( IFN-α/β ) response , the protein kinase R ( PKR ) -mediated stress response is being recognized as an important innate response pathway . Upon detecting viral dsRNA , PKR phosphorylates eIF2α , leading to the inhibition of cellular and viral translation and the formation of stress granules ( SGs ) , which are increasingly recognized as platforms for antiviral signaling pathways . It is unknown whether cellular infection by MERS-CoV activates the stress response pathway or whether the virus has evolved strategies to suppress this infection-limiting pathway . Here , we show that cellular infection with MERS-CoV does not lead to the formation of SGs . By transiently expressing the MERS-CoV accessory proteins individually , we identified a role of protein 4a ( p4a ) in preventing activation of the stress response pathway . Expression of MERS-CoV p4a impeded dsRNA-mediated PKR activation , thereby rescuing translation inhibition and preventing SG formation . In contrast , p4a failed to suppress stress response pathway activation that is independent of PKR and dsRNA . MERS-CoV p4a is a dsRNA binding protein . Mutation of the dsRNA binding motif in p4a disrupted its PKR antagonistic activity . By inserting p4a in a picornavirus lacking its natural PKR antagonist , we showed that p4a exerts PKR antagonistic activity also under infection conditions . However , a recombinant MERS-CoV deficient in p4a expression still suppressed SG formation , indicating the expression of at least one other stress response antagonist . This virus also suppressed the dsRNA-independent stress response pathway . Thus , MERS-CoV interferes with antiviral stress responses using at least two different mechanisms , with p4a suppressing the PKR-dependent stress response pathway , probably by sequestering dsRNA . MERS-CoV p4a represents the first coronavirus stress response antagonist described .
Innate antiviral responses represent the first line of defense against invading viral pathogens . Host cells are equipped with multiple mechanisms to detect and respond to non-self , pathogen-associated molecular patterns ( PAMPs ) . One of these PAMPs , viral cytosolic RNA , can be detected by RIG-I-like receptors ( RLRs ) , such as melanoma differentiation-associated protein 5 ( MDA5 ) and retinoic acid inducible gene 1 ( RIG-I ) . Upon recognition of viral , non-self RNA , signal transduction pathways are activated , which results in the expression of type I interferons ( IFN-α/β ) , proinflammatory cytokines and chemokines . Secreted IFN-α/β triggers the transcription of interferon-stimulated genes ( ISGs ) , both in infected as neighboring cells , and thereby implements an antiviral state that restricts virus propagation in the host . Growing evidence points to an important role of the stress response pathway as an additional innate antiviral response [1 , 2] . One of the ISGs , protein kinase R ( PKR ) , detects viral RNA in the cytoplasm , which induces its autophosphorylation and subsequent phosphorylation of the alpha subunit of eukaryotic translation initiation factor 2 ( eIF2α ) . PKR mediated phosphorylation of eIF2α inactivates ( viral ) protein synthesis , thereby affecting virus propagation . Stalled translation initiation complexes , together with nucleating factors like G3BP1 , G3BP2 , TIA-1 and many translation initiation factors like eIF3 , form cytoplasmic aggregates , which are called stress granules ( SGs ) . The role of these SGs remains controversial , but growing evidence points to a role of these SGs as a platform for antiviral signal transduction [3–5] . To ensure efficient virus replication , many viruses encode proteins with specialized functions to evade innate antiviral responses , although their mode of action and the point of interference may differ . Viruses usually interfere in several antiviral pathways and even disrupt pathways at multiple levels , to ensure efficient suppression of the host innate antiviral responses . A well-studied example is the Influenza A virus NS1 protein , which , among many other evasive functions , shields viral double-stranded RNA ( dsRNA ) from detection by both RLRs and PKR [6 , 7] , thus blocking IFN-α/β and antiviral stress response pathways , respectively . Coronaviruses are large positive-stranded RNA viruses belonging to the order Nidovirales . The coronavirus genome is typically between 26 and 32 kb in size and encodes more than 20 proteins . The 5’ open reading frame ( ORF ) 1ab encodes the non-structural proteins ( nsps ) , which together form the replication-transcription machinery . The 3’ end of the coronavirus genome contains several additional ORFs encoding the structural proteins and a varying number of accessory proteins . These accessory proteins often lack any detectable homology to other viral or host proteins and their function is unknown in many cases . A common feature , however , is that they are often not essential for virus replication per se but are important for virulence , suggesting that accessory proteins serve to modulate host antiviral responses [8–13] . Human coronaviruses generally cause mild respiratory symptoms . Exceptions are severe acute respiratory coronavirus ( SARS-CoV ) , which emerged in China in 2002 through cross-species transmissions from bats and civet cats [14] , and Middle East respiratory syndrome coronavirus ( MERS-CoV ) , which emerged in the Arabian Peninsula in 2012 . MERS-CoV causes acute and severe respiratory symptoms and continues to make a serious impact on the local as well as the global health system with over 1 , 694 laboratory confirmed cases and 605 deaths as of March 21st 2016 [15] . This virus is believed to be transmitted to humans primarily via animal hosts , most likely dromedary camels [16 , 17] . As yet , little is known about how MERS-CoV modulates host antiviral responses . There is firm evidence that MERS-CoV inhibits IFN-α/β production [18–20] and several viral proteins have been implicated in this evasion mechanism–including accessory protein 4a ( p4a ) , which is a dsRNA-binding protein [21–23]–but the inhibitory effect of these proteins on innate antiviral responses has thus far only been demonstrated in transfected cells expressing these viral proteins , not during virus infection . Whether MERS-CoV has also evolved mechanisms to modulate the stress response pathway is unknown thus far . Here , we show for the first time that MERS-CoV actively suppresses the stress response pathway and we identify the accessory protein 4a as a potent inhibitor of the PKR-mediated stress response pathway . Furthermore , we provide evidence that the rescue of translation and inhibition of SG formation rely on p4a’s dsRNA-binding function , suggesting that it exerts antagonistic activity by sequestering dsRNA from recognition by PKR . Moreover , evidence for the existence of at least one other MERS-CoV encoded stress response antagonist is provided .
To investigate whether MERS-CoV infection activates the stress response pathway , Vero cells were infected with MERS-CoV ( MOI = 1 ) and analyzed for the occurrence of SG at regular time intervals by visualizing the subcellular localization of eIF3 and G3BP2 , which are established markers for SGs . In parallel , the efficiency of virus infection was monitored by visualizing dsRNA using the J2 antibody . Despite efficient virus infection and replication , as indicated by the accumulation of considerable amounts of viral dsRNA in the cytosol , no SGs were observed at any of the indicated time points ( Fig 1A ) . The lack of SGs was not due to an intrinsic defect in the stress response pathway of Vero cells as clear SGs were formed upon arsenic acid treatment and poly ( I:C ) transfection ( Fig 1B ) . Together , these findings indicate that MERS-CoV either hides its viral RNA from detection by PKR , possibly through the formation of double membrane vesicles [24] , and/or that it encodes one or more antagonists to suppress activation of the stress response pathway . To investigate whether MERS-CoV accessory proteins can suppress the stress response pathway , we expressed them individually as EGFP fusion proteins and monitored SG formation in transfected cells . This approach is based on the observation that transfection of plasmid DNA , and in particular the pEGFP plasmids , can activate PKR , most likely due to the production of dsRNA formed from positive and negative sense mRNA transcription from cryptic promoters in these plasmids [25] . Indeed , we observed that transfection of pEGFP plasmid DNA in HeLa cells triggered SG formation in a PKR-dependent manner , as no SGs were observed in PKR knockout cells ( HeLa-PKRKO ) , which we generated using the CRISPR-Cas9 system ( S1 Fig ) ( Fig 2A and 2B ) . Also , using the J2 anti-dsRNA antibody , we noticed a significant increase in dsRNA levels in cells transfected with pEGFP plasmid DNA and especially in cells that displayed SGs ( Fig 2C and 2D ) . This phenomenon was not restricted to the pEGFP plasmid as all plasmids with eukaryotic promoters induced SG formation in our HeLa cells , albeit to different levels , while those with prokaryotic promoters did not ( S2 Fig ) . Together , these data support the idea that transfection of pEGFP plasmid DNA can trigger dsRNA-dependent and PKR-mediated SG formation , and provide the basis for a convenient and versatile method to test potential antagonistic activities of viral proteins by expressing them as EGFP fusion proteins . Plasmids each encoding one of the four MERS-CoV accessory proteins fused to EGFP were transfected into HeLa cells . As a positive control , we took along an EGFP fusion of the influenza A virus ( IAV ) NS1 protein , which is an established PKR antagonist . As shown in Fig 2E , plasmid DNA transfection induced SG formation except for the plasmids encoding the MERS-CoV p4a and IAV NS1 EGFP fusion proteins . The absence of SG formation ( Fig 2E and 2F ) coincided with a lack of PKR phosphorylation ( Fig 2G ) . We also tested the ability of these MERS-CoV accessory proteins to suppress the stress response pathway induced by the more commonly applied method of poly ( I:C ) transfection . Again , we observed that p4a , but none of the other MERS-CoV accessory proteins , suppressed SG formation ( S3 Fig ) . The inhibitory effect of p4a , as well as that of NS1 , was less pronounced in this assay , possibly because the relatively large amounts of poly ( I:C ) may exceed the maximum capacity of the PKR antagonists . Taken together , our data suggests that MERS-CoV p4a is a PKR antagonist and inhibits the stress response pathway at the level of , or upstream of , PKR phosphorylation . We observed that the protein levels of p4a and NS1 were higher than those of the other MERS-CoV accessory proteins ( Fig 2E ) . We reasoned that the inhibition of plasmid DNA-induced PKR activation may increase protein translation levels . Indeed , co-expression of p4a or NS1 together with Renilla luciferase ( RLuc ) caused a reproducible 5- to 10-fold increase in luciferase counts compared to the EGFP control plasmid ( Fig 3A ) . This effect was attributed to increased translation , since p4a expression had no effect on RLuc mRNA levels . In addition , RLuc counts were not increased in PKRKO cells , indicating that p4a increases translation efficiency via inhibition of PKR ( S4 Fig ) . Other established viral PKR antagonists like the Vaccinia virus E3L [26] and Ebola virus VP35 [27] caused a similar increase in RLuc expression levels . Comparable results were obtained upon co-expression with an RFP expression plasmid ( Fig 3B ) . These data are in line with the observation that MERS-CoV p4a antagonizes PKR activity , and provide another indication that viral PKR antagonists can rescue translation efficiency in cells in which the stress pathway is activated by ( viral ) dsRNA . Both MERS-CoV p4a and IAV NS1 are dsRNA binding proteins [6 , 21] , which suggests that p4a shields the viral dsRNA from detection by PKR . To test whether p4a can also inhibit stress pathway activation via PKR- and dsRNA-independent mechanisms , we used arsenic acid and heat shock to induce eIF2α-dependent stress pathway activation [28] . Furthermore , we used pateamine A to induce SG formation via an eIF2α-independent mechanism [29] . In agreement with earlier findings , IAV NS1 failed to inhibit PKR-independent SG formation [30] . A small reduction in PKR-independent SG formation was observed in cells overexpressing p4a ( Fig 4A and 4B ) . However , lack of SGs was only observed in cells expressing very high levels of p4a , whereas a moderate expression level of p4a was already sufficient to inhibit PKR-mediated SG formation ( Fig 2E ) . To rule out any involvement of PKR expression in the small reduction of PKR-independent SG formation , we tested arsenic acid , heat shock and pateamine A-induced stress pathway activation in HeLa-PKRKO cells . Also under these conditions , expression of p4a affected SG formation only in a small fraction of the cells ( Fig 4C and 4D ) . Thus , MERS-CoV p4a seems to predominantly suppress dsRNA-dependent PKR activation and does not efficiently target other parts of the stress response pathway . Studying immune evasion functions of viral proteins by transient overexpression from plasmid DNA may suffer from shortcomings . Transfection procedures fail to mimic the dynamic interplay between dsRNA and the antagonist , both of which gradually appear over time during the viral life cycle . Furthermore , transfection may yield non-physiologically high levels of viral proteins and/or dsRNA mimics which may blur results . Also , dsRNA-mimicking molecules , like poly ( I:C ) , may be delivered to compartments where viral dsRNA does not naturally localize under infection conditions . Therefore , we set out to investigate the function of p4a as an innate antiviral response antagonist under infection conditions . For this , we made use of a recombinant encephalomyocarditis virus ( EMCV , strain mengovirus ) . EMCV is a member of the picornavirus family that , like coronaviruses , produces dsRNA replication intermediates during its life cycle . In the recombinant EMCV , the function of the leader ( L ) protein–which antagonizes the dsRNA-triggered IFN-α/β and stress response pathways–is disturbed by specific mutations in an essential zinc-finger motif ( EMCV-L-Zn ) [31 , 32] . By consequence , and in contrast to wt virus , EMCV-L-Zn causes strong activation of the IFN-α/β and stress response pathways [31 , 32] . To study whether heterologous expression of p4a can prevent PKR activation , recombinant viruses were generated expressing Strep2-tagged MERS-CoV p4a or IAV NS1 ( as a control ) upstream of the inactivated L ( Fig 5A ) . EMCV wt infection did not induce SG formation while EMCV-L-Zn induced SGs in ~80% of the cells . Infection of cells with recombinant EMCV-L-Zn expressing p4a or NS1 protein resulted in SG formation in <20% of the cells ( Fig 5B and 5C ) . This reduction was not due to differences in infection efficiency , since Strep2-tagged proteins were detected in the majority of cells ( Fig 4B ) . In fact , SGs were only observed in cells displaying low expression levels of p4a or IAV NS1 . Western blot analysis was performed to assess the level of PKR phosphorylation . Total PKR levels were significantly reduced in EMCV-infected cells , a phenomenon that was described earlier by Dubois et al . [33] , although the mechanism behind this remains unclear . Yet , even with these reduced PKR levels , EMCV-L-Zn infection induced strong PKR phosphorylation , which was reversed by the expression of p4a or NS1 ( Fig 5D ) . Analysis of viral protein levels using an antibody directed against the viral capsid indicated that viral protein levels were higher in cells infected with p4a- and NS1-expressing viruses compared to EMCV-L-Zn infected cells , indicating that expression of these PKR antagonists increased virus replication efficiency . Taken together , these results indicate that MERS-CoV p4a can functionally replace the PKR antagonist of a picornavirus in infected cells . MERS-CoV p4a contains a dsRNA-binding motif similar to those found in some cellular proteins ( S5 Fig ) . Previously , a p4a mutant containing substitutions in its dsRNA-binding motif ( K63A/K67A ) was shown to be deficient in binding dsRNA [22] . Based on the sequence similarity of this dsRNA-binding motif to those in Staufen , ADAR1 , ADAR2 and PKR , and the published NMR structure of the ADAR2 dsRNA-binding domain in complex with its ligand [34] , we designed a second mutant containing a single substitution ( Q9P ) in another part of the conserved dsRNA-binding motif ( S5 Fig ) . Infection of HeLa cells with recombinant EMCV-L-Zn viruses expressing either of these p4a mutants resulted in efficient SG formation , indicating a complete loss of the stress-antagonizing function ( Fig 6A and 6B ) . In agreement herewith , analysis of the PKR phosphorylation status demonstrated that the p4a mutants failed to inhibit PKR phosphorylation ( Fig 6C ) . Consistently , viruses expressing these mutants showed reduced capsid protein expression , possibly as a consequence of PKR-mediated translation inhibition . Thus , the dsRNA-binding motif in MERS-CoV p4a is essential for its function to antagonize PKR-mediated SG formation and translation shut-off . Previous studies have shown that expression of p4a is able to reduce the level of IFN-α/β pathway activation in transiently transfected cells [21–23] . Consistently , we observed that transient expression of p4a inhibited poly ( I:C ) -induced ( Fig 7A ) and dsRNA-induced ( Fig 7B ) IFNβ mRNA transcription . To assess whether p4a can also inhibit the IFN-α/β pathway in virus-infected cells , we compared IFNβ mRNA transcription levels in cells infected with recombinant EMCV-L-Zn viruses expressing either p4a or NS1 . Both p4a and IAV NS1 significantly suppressed transcription of IFNβ mRNA ( Fig 7C ) . This ability was lost in viruses expressing mutant p4a proteins that are unable to bind dsRNA ( Fig 7 ) . These data show that MERS-CoV p4a also inhibits the IFN-α/β response in EMCV-L-Zn-infected cells and that this function also requires its dsRNA-binding activity . Our data show that p4a is a multi-functional protein that antagonizes both the stress response and the IFN-α/β response pathways . To demonstrate the functional and beneficial role of p4a-mediated antagonism of the stress response pathway , we set out to compare the replication efficiency of recombinant viruses in HeLa-wt cells and cells that are defective in the PKR-induced stress response pathway ( HeLa-PKRKO cells ) . Infection of HeLa-PKRKO cells with EMCV-L-Zn showed that these cells are unable to mount a stress response ( Fig 8A ) , whereas IFN-α/β pathway activation was only slightly affected in these cells ( Fig 8B ) , indicating that possible differences in virus fitness can be predominantly attributed to the defective stress response pathway . Replication of EMCV-L-Zn under low MOI infection conditions is severely impaired in HeLa-wt cells , whereas replication was fully rescued to the level of EMCV wt in HeLa-PKRKO cells ( Fig 8C ) . Comparison of the replication efficiency of recombinant viruses expressing p4a or the p4a mutant containing the K63A/K67A substitutions showed that the antagonistic activity of p4a provided a clear fitness advantage in HeLa-wt cells ( Fig 8C ) . The observation that the p4a-expressing virus failed to replicate to similar titers as wt virus is unlikely due to inefficient PKR inhibition by p4a as comparable titers were obtained for the recombinant viruses expressing p4a or mutant p4a in HeLa-PKRKO cells . Notwithstanding the lower virus titer , which may either be due to imperfect polyprotein processing due to introduction of p4a or to less efficient encapsidation of the larger viral genome , these results provide evidence that the PKR antagonistic function of MERS-CoV p4a can provide a virus fitness advantage in PKR-competent cells . Similar results were obtained in virus competition experiments ( Fig 8D ) , which is a more sensitive method to compare virus fitness and can reveal smaller fitness differences . Upon low MOI infection , EMCV-L-Zn expressing p4a rapidly outgrew EMCV-L-Zn in HeLa-wt cells but not in HeLa-PKRKO cells ( Fig 8E ) . No fitness advantage was observed with virus expressing the mutant p4a ( Fig 8F ) . We also co-infected cells with viruses expressing either p4a or mutant p4a . Since these viruses could not be distinguished based on their amplicon length , we used a HindIII restriction reaction to specifically cleave the wt 4a PCR fragment ( the HindIII site is absent in the mutant 4a gene ) . Consistent with the results of the multi-cycle infection experiment shown in Fig 8C , the virus expressing p4a replicated better than the virus expressing mutant p4a in HeLa-wt cells whereas in PKRKO cells only a minor advantage was observed ( Fig 8G ) . Thus far , we used a recombinant picornavirus , EMCV , to analyze the function of p4a in virus-infected cells , in the absence of other MERS-CoV proteins . To assess the relevance of p4a for stress response antagonism in MERS-CoV infected cells , we used recombinant MERS-CoVΔORF4 that is deficient in p4a and p4b expression . Surprisingly , like wt MERS-CoV , MERS-CoVΔORF4 did not induce SG formation in Vero cells ( Fig 9A ) , suggesting that MERS-CoV expresses at least one other protein that suppresses the stress response pathway . To gain more insight into the working mechanism of this other stress response pathway antagonist , we treated MERS-CoV infected cells with arsenite . As demonstrated in Fig 9B , this treatment resulted in SG formation in all the uninfected cells , whereas no SGs were detected in cells infected with either MERS-CoV or MERS-CoVΔORF4 ( Fig 9B ) . These findings strongly suggest that MERS-CoV encodes at least one other stress response antagonist with a mode of action that differs from that of p4a . We also tested the IFN-α/β pathway activation in cells infected with the mutant virus . In line with the reports that several MERS-CoV proteins can antagonize the IFN-α/β pathway [21 , 23 , 35–38] , no increase in IFNβ mRNA levels was observed in Huh7 cells infected with MERS-CoV or MERS-CoVΔORF4 ( Fig 9C ) . Taken together , these data provide evidence for substantial redundancy with respect to antagonism of innate antiviral responses in MERS-CoV infected cells .
Most viruses have evolved mechanisms to antagonize innate antiviral responses . Coronaviruses encode a set of genus-specific , or in some cases even species-specific , proteins that are generally dispensable for replication in vitro but ensure efficient virus replication and/or spreading in vivo [10 , 11 , 39–41] . Some of these so-called accessory proteins have been shown to antagonize specific innate antiviral responses , but the functions of most of them are still unknown [9 , 10 , 23 , 42–44] . Thus far , most studies concentrated on IFN-α/β pathway antagonists , whereas inhibition of the cellular stress response pathway by coronaviruses remains largely unexplored . In this study , we focused on the recently identified MERS-CoV , and showed that infected cells fail to activate the stress response pathway . In our subsequent search for MERS-CoV-encoded stress response antagonists , each of its accessory proteins was tested individually for the ability to suppress this pathway . Transient expression of p4a specifically suppressed dsRNA-mediated and PKR-dependent translation inhibition and SG formation . Moreover , we showed that p4a can functionally substitute for the PKR antagonist of EMCV in infected cells . Introduction of specific mutations revealed that the ability of p4a to suppress activation of the stress response pathway depends on its dsRNA-binding function . Together , the data strongly suggest that p4a suppresses the PKR-mediated stress response pathway by sequestering viral dsRNA . Yet , infection of cells with a recombinant MERS-CoV deficient in p4a expression failed to trigger SG formation . This finding points to the expression of at least one other stress response antagonist by MERS-CoV . Importantly , this other suppressor ( s ) differs in its mode of action of p4a , since in contrast to p4a , it was able to suppress activation of the arsenite-induced stress pathway . Together , these data suggest that MERS-CoV has evolved redundant mechanisms to suppress the stress response pathway at multiple levels . To our knowledge , MERS-CoV p4a is the first coronavirus protein identified as an antagonist of the dsRNA-dependent , PKR-mediated stress response . There are strong indications that other coronaviruses also encode stress response antagonists but their identity and mode of action remain to be determined . Infectious bronchitis virus ( IBV , a γ-CoV ) interferes with phosphorylation of both PKR and eIF2α through an unknown mechanism ( s ) [45] . Transmissible gastroenteritis virus ( TGEV ) Transmissible gastroenteritis virus ( TGEV ) triggers SG formation , but causes a reduction in the amount of phosphorylated eIF2α over time , possibly by recruiting eIF2α phosphatase PP1 through accessory protein 7 [46 , 47] . Mouse hepatitis virus ( MHV , a lineage A β-CoV ) triggers eIF2α phosphorylation and SG formation relatively late in infection , suggesting that the virus actively delays the stress response pathway [48–50] , but the mechanism is unknown . SARS-CoV ( a lineage B β-CoV ) has been reported to trigger PKR activation but to be resistant to its antiviral activity [51] , although in another study a strong antiviral effect of PKR was observed [52] . Hence , the limited information that is available suggests that coronaviruses have acquired different strategies to antagonize the stress response pathway . Importantly , none of these coronaviruses encode a protein with any homology to MERS-CoV p4a . In this study , we assessed p4a’s antagonistic activities not only upon transient overexpression , but also in the context of viral infection . For this , we introduced p4a in a recombinant EMCV ( EMCV-L-Zn ) in which the IFN-α/β and stress response pathway antagonist—the leader ( L ) protein—was inactivated . A p4a-expressing recombinant EMCV may provide several advantages over overexpression through transient transfection , as it likely better mimics the dynamic production of—as well as the interplay between—dsRNA and the viral antagonist . Using this approach , we showed that dsRNA sequestration by p4a efficiently suppresses the PKR-dependent stress response pathway as well as MDA5-mediated IFN-α/β responses under these infection conditions , and thereby provides a fitness advantage to this recombinant EMCV . Similar results were obtained with a recombinant virus expressing IAV NS1 , which was included as a control . Together , these data suggests that p4a can be categorized in the group of previously identified viral dsRNA-binding antagonists of stress response and IFN-α/β pathways , which besides IAV NS1 , also includes Ebola virus VP35 and Vaccinia virus E3L [26 , 27] . Our results showed that besides p4a , MERS-CoV expresses at least one other stress response antagonist . This other antagonist ( s ) is likely one of the nsps or a structural protein , as we excluded stress-antagonizing roles of the other accessory proteins . At least one antagonist can also suppress the arsenite-induced stress response pathway , and is therefore unlikely to act directly at the level of PKR . Instead , it may act at the level of eIF2α phosphorylation or SG formation . Identification of the other stress response antagonist ( s ) and elucidation of its/their mode of action , awaits further investigation . Functional redundancy in suppressing innate antiviral responses is a well-documented phenomenon for coronaviruses . The MERS-CoV accessory proteins ( p4a , p4b , and p5 ) [21–23 , 36] as well as the structural M protein and the ORF1ab-encoded nsp3 [38 , 37] , have all been implicated in antagonizing IFN-α/β pathway activation . This provides a likely explanation for our observation that recombinant MERS-CoV lacking p4a and p4b was still able to suppress IFN-β mRNA transcription . MERS-CoV p4a homologs have exclusively been identified in lineage C β-CoVs , which besides MERS-CoV comprises a MERS-like coronavirus found in European hedgehogs [53] , and bat coronaviruses ( BatCoV ) HKU4 and HKU5 [54–56] . The p4a-like accessory proteins of these other lineage C viruses all contain dsRNA-binding motifs and may therefore have similar functions as MERS-CoV p4a . Yet , a study by Siu et al . indicated that p4a of BatCoV-HKU4 , in contrast to that of MERS-CoV and BatCoV-HKU5 , does not bind poly ( I:C ) and does not inhibit IFN-α/β responses [22] . If BatCoV-HKU4 p4a is indeed unable to sequester dsRNA , then it is likely unable to suppress the dsRNA-triggered stress response pathway as well . Interestingly , sequence analysis of a MERS-CoV strain isolated from patients in Jordan identified a 16 amino acid deletion in p4a [57] . This deletion does not affect the residues comprising the dsRNA binding site . However , as it removes the second β-strand in the classical αβββα-fold of the dsRNA binding domain , p4a’s dsRNA binding properties and , in consequence , its function as antagonist , are most likely compromised . If so , stress antagonism by p4a may be dispensable for MERS-CoV replication and transmission among humans . Increasing evidence suggests that coronavirus accessory proteins often have niche-specific ( e . g . organ- or tissue-specific ) or host-tailored functions . For example , accessory protein 3c is required for replication of low-virulence feline enteric coronavirus ( FECV ) , which primarily replicates in the enteric tract , but not for replication of FECV-derived , highly virulent feline infectious peritonitis virus ( FIPV ) isolates , which have acquired the ability to replicate in macrophages [58 , 59] . Also , accessory proteins contributing to viral fitness in one particular host species may sometimes prove less important in a novel host following a species-jump . For example , in SARS-CoV and CoV-229E some accessory genes were lost through gradual deletion following the introduction of these viruses into humans [60 , 61] . Acquisition as well as loss of accessory proteins may reflect adaptations to different immunological environments in different niches or hosts . In this study , we showed that MERS-CoV p4a can potently antagonize innate antiviral responses in human cells . Yet , as suggested by the Jordan outbreak , p4a may not be critical for zoonotic transmission nor for limited human-to-human spread , possibly because of redundancy in viral anti-stress response strategies . Whether p4a will be lost or maintained in the hapless event MERS-CoV establishes sustained community transmission remains an open question .
HeLa-R19 , Huh7 and BHK-21 cells were maintained in Dulbecco’s Modified Eagle’s Medium ( DMEM ) supplemented with 10% ( V/V ) fetal calf serum ( FCS ) . Vero cells ( ATCC CCL-81 ) were grown in Eagle’s minimum essential medium with 8% FCS , 100 units/ml penicillin and streptomycin , 2 mM L-glutamine and non-essential amino acids . MERS-CoV infections [62] were carried out as described previously [24 , 63] inside biosafety cabinets in BSL III facilities at Leiden University Medical Center and Universidad Autonoma de Madrid . Recombinant MERS-CoVs that were used in Madrid have been described previously [63] . Recombinant MERS-CoVs that were used in Leiden were derived from the previously described infectious MERS-CoV clone pBAC-MERSFL [63] , and adapted as follows using two step en-passant in vivo recombineering reactions in E . coli [64] . The CMV promoter at the 5’end of the MERS-CoV cDNA sequence was replaced by a T7 RNA polymerase promoter and a unique NotI linearization site was inserted at the 3’end , so that the virus could be launched from transfecting in vitro synthesized RNA transcripts ( produced using an mMESSAGE mMACHINE T7 transcription kit from ThermoFisher scientific ) . To construct MERS-CoVΔORF4 from this adapted clone , the coding sequence of MERS-CoV p4a/p4b was removed and replaced by a red fluorescent protein ( RFP ) gene , which however for unclear reasons did not result in red fluorescence during infection . All the genetic modifications to the original pBAC-MERSFL were verified by sequencing . The MERS-CoVΔORF4 virus grew to similar titers as the recombinant wt MERS-CoV derived from the original clone . Recombinant EMCV viruses were derived from the pM16 . 1 infectious clone [65] . The pStrep2-VFETQG-Zn-M16 . 1 infectious clone was constructed using site-directed mutagenesis ( SDM ) using the pCVB3-3Cpro-QG-M16 . 1 as template DNA [32] . The Zn-finger mutation in L was introduced by SDM using the following oligonucleotides: Fw; 5’-ATGACCTTTGAAGAAGCCCCAAAAGCCTCCGCCTTACAATAC-3’ and Rv; 5’- GGAATGAGCACAAATCTCTTG-3’ . The optimized 3Cpro recognition site ( VFETQG ) was introduced by SDM using the following oligonucleotides: Fw; 5’-GAAACTCAAGGCGCAACGACTATGGAGC-3’ and Rv; 5’-AAAGACCGCGGCCGCTTGCTCATCATTG-3’ . Finally , the Strep2-tag was introduced by SDM using the following oligonucleotides: Fw; 5’-GGCCGCCTGGTCACATCCTCAGTTTGAGAAGGGTGCCTGGTCTCATCCCCAATTCGAAAA-3’ and Rv: 5’- GGCCTTTTCGAATTGGGGATGAGACCAGGCACCCTTCTCAAACTGAGGATGTGACCAGGC-3’ . Genes of interest were inserted into the XhoI/NotI restriction sites of the pStrep2-VFETQG-Zn-M16 . 1 infectious clone . Viruses were recovered by transfection of run-off RNA transcripts into BHK-21 cells . Upon total CPE , cells were subjected to three freeze-thaw cycles and cell debris was pellet at 4 , 000xg for 15 minutes . Virus was concentrated by ultracentrifugation though a 30% sucrose cushion at 140 , 000xg for 16 hours in a SW32Ti rotor . HeLa-R19 PKRKO were generated using the CRISPR/Cas9 system as previously described [66] . Briefly , gRNA encoding oligonucleotides cassettes to target human PKR ( gRNA1: 5’-ACCGGACCTCCACATGATAGG-3’ and 5’-AACCCTATCATGTGGAGGTCC-3’ , gRNA2: 5’-CCGTACTACTCCCTGCTTCTGAG-3’ and 5’-AAACTCAGAAGCAGGGAGTAGTA-3’ ) were cloned into the SapI restriction sites of the pCRISPR-hCas9-2xgRNA-Puro plasmid . HeLa-R19 cells were seeded in 6-well clusters ( 100 , 000 cells/well ) and next day transfected with 2 . 5 μg plasmid DNA using Fugene6 ( Promega ) according to manufacturer’s instructions . Next day successfully transfected cells were selected using puromycin and single-cell clones were generated using end-point dilutions . Knockout efficiency was determined by sequence analysis of the PKR locus in the genomic DNA and western blot analysis ( S1 Fig ) . Arsenic acid was purchased at Sigma-Aldrich and used at a final concentration of 0 . 5 mM in DMEM . Pateamine A was kindly provided by Prof . Jerry Pelletier [67] and used at a concentration of 100 nM in DMEM . Poly ( I:C ) was purchased from GE Healthcare and dsRNA ligand was prepared using the Replicator RNAi kit ( Finnzymes ) using the following oligonucleotides ( Fw , possessing T7 promoter sequence ) TAATACGACTCACTATAGGGGATACAGTGACAGGGCG and ( Rv , possessing Phi6 promoter sequence ) GGAAAAAAACCGCACCGAATGCGGAGAATTTAC and the pRib-CVB3/T7 Coxsackie virus B3 infectious clone as template [68] . Expression plasmids encoding enhanced green fluorescent protein ( EGFP ) tagged proteins were created by PCR amplification of the gene of interest with oligonucleotides flanked by XhoI ( Fw ) or BamHI ( Rv ) restriction sites ( MERS-CoV ORF3: 5’-AAAAACTCGAGATGAGAGTTCAAAGACCACCC-3’ and 5’-AAAAAGGATCCATTAACTGAGTAACCAACGTCAAAAAG-3’ , ORF4a: 5’-AAAAACTCGAGATG GATTACGTGTCTCTGCTTAATC-3’ and 5’-AAAAAGGATCCGTGGGAGAATGGCTCCTC-3’ , ORF4b: 5’-AAAAACTCGAGATGGAGGAATCCCTGATGGATG-3’ and 5’-AAAAAGGATCCAAA TCCTGGATGATGTAAAATGGGG-3’ , ORF5: 5’-AAAAACTCGAGATGGCTTTCTCGGCGTC-3’ and 5’-AAAAAGGATCCAACGATAAGCGAGCTCGG-3’ , IAV NS1: 5’-AAAAACTCGAGATGGAT CCAAACACTGTGTC-3’ and 5’-AAAAAGGATCCAACTTCTGACCTAATTGTTC-3’ , VV E3L: 5’-AAAAACTCGAGATGTCTAAGATCTATATTGACGAGCGTTCTG-3’ and 5’-AAAAAGGATCCG AATCTAATGATGACGTAACCAAGAAGTTTATCTACTG-3’ , Ebola VP35: 5’-AAAAACTCGAGATGAC AACTAGAACAAAGGGCAGGG-3’ and 5’-AAAAAGGATCCAATTTTGAGTCCAAGTGTTTTACC ATCTTGAAGC-3’ . Digested PCR products were ligates into XhoI/BamHI digested pEGFP-N3 plasmid and gene integrity was confirmed by sequencing analysis . pcDNA-RFP expression plasmid was constructed by PCR amplification of the RFP gene using oligonucleotides flanked by NheI ( Fw ) and NotI ( Rv ) restriction sites ( Fw ) GCTAGCGCCACAACCATGGCCTCCTCCGAGGAC and ( Rv ) GCGGCCGCCGGCGCCGGTGGAGTGGCGGCCCTC and subsequently cloning into the NheI/NotI digested pcDNA-EGFP plasmid [69] . The pJET-puro ( puromycin resistance vector ) plasmid was developed by ligation of the EF1a-Puro fragment into the pJet1 . 2/blunt vector ( Thermo Fisher ) . pRL-TK ( Renilla luc expression vector ) plasmid was purchased from Promega . HeLa-R19 cells were seeded in a 96-wells cluster ( 5 , 000 cells/well ) and the next day they were transfected with the indicated plasmids ( 40 ng pEGFP , 10 ng pRL-TK ) using Fugene6 . 24 hours post transfection , cells were lysed in 20 μl passive lysis buffer ( Promega ) and analyzed on the Centro LB 960 Microplate Luminometer ( Berthold technologies ) using the Renilla luciferase reporter kit ( Promega ) according to manufacturer instructions . Cells were seeded in a 24-wells cluster ( 50 , 000 cells/well ) and the next day they were transfected with the indicated plasmids ( 500 ng/well; 250 ng/plasmid ) using Fugene6 . Twenty-four hours post transfection , cells were released using trypsin , washed once in phosphate buffered saline ( PBS ) and fixed for 30 minutes with 2% paraformaldehyde ( PFA ) in PBS . Cells were analyzed on FACS Canto ( BD ) using BD FACS Diva software . Cells were seeded on glass slides in a 24 wells cluster ( 25 , 000 cells/well ) and the next day they were infected ( MOI = 10 ) or transfected ( 500 ng total DNA ) using Fugene6 . At 6h post infection or 24h post transfection , cells were fixed using 4% PFA in PBS for 30 minutes at RT . Vero cells seeded on glass slides were transfected with 1 μg Poly ( I:C ) per 6-well using Lipofectamine 2000 ( Thermo Fisher Scientific ) . Cells were permeabilized with PBS + 0 . 2% Triton X-100 , washed trice with blocking buffer ( PBS + 2% bovine serum albumin [BSA] + 50mM NH4Cl ) , and incubated with blocking buffer for 1 h . Cell monolayers were incubated for 1 h with primary antibody mouse-α-G3BP1 ( BD , 1:1 , 000 ) , rabbit-α-TIA1 ( Santa-Cruz , 1:50 ) , mouse-α-dsRNA ( J2 , English&Scientific Consulting , 1:1 , 000 ) , goat-α-eIF3 ( Santa-Cruz , 1:100 ) , rabbit-α-G3BP2 ( Bethyl Laboratory , 1:200; or Assay Biotech , 1:500 ) , or rabbit-α-MERS-CoV ( 1:500: raised against the MERS-CoV M carboxyl terminal peptide CRYKAGNYRSPPITADIELALLRA ) , and then for 30 min with secondary antibody donkey-α-mouse-Cy3 ( Jackson ImmunoResearch , 1:1000 ) , donkey-α-rabbit-Alexa488 ( Jackson ImmunoResearch , 1:1000 ) , bovine-α-goat-Alexa647 ( Jackson ImmunoResearch , 1:1000 ) , donkey-α-rabbit-Cy5 ( Jackson ImmunoResearch , 1:200 ) , donkey-α-mouse-Alexa 488 ( Invitrogen , 1:200 ) or donkey-α-goat-Alexa 594 ( Invitrogen , 1:200 ) and Hoechst-33258 ( 1:2 , 000 ) diluted in blocking buffer . Between and after the incubations , the cell monolayers were washed three times with blocking buffer . Finally , the cells were washed once with distilled water and coverslips were mounted on glass slides in FluorSafe ( Calbiochem ) . Cells were examined by confocal microscopy ( Leica SPE-II ) . Cells were seeded in 10-cm dishes ( 2 . 5 x 106 cells/dish ) and the next day cells were infected ( MOI = 10 ) or transfected ( 8 μg plasmid DNA ) using Fugene6 . At 6h post infection or 24h post transfection , cells were released using trypsin , washed once in wash buffer ( 100 mM Tris/HCl pH 8 , 0 + 1 mM EDTA + 50 mM NaCl ) and lysed in 200 μl lysis buffer ( 100 mM Tris/HCl pH 8 , 0 + 1 mM EDTA + 50 mM NaCl + 1% NP40 + protease inhibitor mix [Roche] + phosphatase inhibitor cocktails #2 and #3 [Sigma-Aldrich] ) . Cell debris was pelleted at 15 , 000 x g for 15 min and 10 μl of cleared cell lysates were resolved using reducing sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) and transferred to 0 . 2 μm nitrocellulose membranes by wet electrophoretic transfer . Membranes were washed once with washing buffer ( PBS + 0 . 1% Tween 20 ) and incubated 1h in blocking buffer ( PBS + 0 . 1% Tween 20 + 2% BSA ) . Membranes were successively incubated for 1 h with primary antibody mouse-α-PKR ( BD , 1:1 , 000 ) , rabbit-α-PKR-P[T446] ( Abcam , 1:2 . 000 ) , mouse-α-Tubulin ( Sigma , 1:5 . 000 ) , rabbit-α-mengovirus capsid ( kindly provided by Prof . Ann Palmenberg , 1:1 . 000 ) or mouse-α-StrepMab classic ( IBA , 1:1 . 000 ) and then for 30 min with goat-α-mouse-IRDye680 ( Li-COR , 1:15 , 000 ) or goat-α-rabbit-IRDye800 ( Li-COR , 1:15 , 000 ) diluted in blocking buffer . Between and after the incubations , the membranes were washed , thrice each time , with washing buffer . Finally , membranes were washed once with PBS and scanned using an Odyssey Imager ( Li-COR ) . RNA isolation , cDNA synthesis , and RT-qPCR were performed as described elsewhere [66 , 63] . | Human coronaviruses generally cause relatively mild respiratory disease . In the past 15 years , the world has witnessed the emergence of two coronaviruses with high mortality rates in humans; severe acute respiratory syndrome coronavirus ( SARS-CoV ) in 2002 and Middle East respiratory syndrome coronavirus ( MERS-CoV ) in 2012 , both originating from animal reservoirs . Successful infection of a host not only depends on the presence of an appropriate receptor but also on the ability of a virus to evade innate antiviral host responses , which constitute the first line of defense against invading viruses . MERS-CoV has been reported to actively suppress the IFN-α/β response , but it is unknown whether it also interferes with another important innate antiviral response , the stress response pathway . Activation of this pathway by a kinase , PKR , curtails virus infection by shutting off cellular and viral protein synthesis . To date , no coronavirus protein has been recognized to suppress the stress response pathway . Here , we show that the accessory protein 4a of MERS-CoV is a potent stress antagonist that prevents PKR activation by sequestering its ligand , dsRNA . This finding furthers our understanding of the molecular mechanism used by MERS-CoV to evade infection-limiting antiviral host responses and may provide new avenues for therapeutic intervention . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"transfection",
"fluorescence",
"imaging",
"medicine",
"and",
"health",
"sciences",
"coronaviruses",
"cellular",
"stress",
"responses",
"pathology",
"and",
"laboratory",
"medicine",
"respiratory",
"infections",
"pathogens",
"cell",
"processes",
"microbiology",
"pulmonology",
"plasmid",
"construction",
"viruses",
"rna",
"viruses",
"dna",
"construction",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"imaging",
"techniques",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"recombinant",
"proteins",
"viral",
"replication",
"molecular",
"biology",
"biochemistry",
"cell",
"biology",
"virology",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"organisms"
]
| 2016 | Middle East Respiratory Coronavirus Accessory Protein 4a Inhibits PKR-Mediated Antiviral Stress Responses |
Foodborne disease outbreaks of recent years demonstrate that due to increasingly interconnected supply chains these type of crisis situations have the potential to affect thousands of people , leading to significant healthcare costs , loss of revenue for food companies , and—in the worst cases—death . When a disease outbreak is detected , identifying the contaminated food quickly is vital to minimize suffering and limit economic losses . Here we present a likelihood-based approach that has the potential to accelerate the time needed to identify possibly contaminated food products , which is based on exploitation of food products sales data and the distribution of foodborne illness case reports . Using a real world food sales data set and artificially generated outbreak scenarios , we show that this method performs very well for contamination scenarios originating from a single “guilty” food product . As it is neither always possible nor necessary to identify the single offending product , the method has been extended such that it can be used as a binary classifier . With this extension it is possible to generate a set of potentially “guilty” products that contains the real outbreak source with very high accuracy . Furthermore we explore the patterns of food distributions that lead to “hard-to-identify” foods , the possibility of identifying these food groups a priori , and the extent to which the likelihood-based method can be used to quantify uncertainty . We find that high spatial correlation of sales data between products may be a useful indicator for “hard-to-identify” products .
In recent years global trade has significantly altered the topology of food supply chains [1] . As a result , the potential impact of contamination events has increased [2] . Worldwide , foodborne illness causes billions of dollars in healthcare related costs each year [3] , and more in economic losses to farmers , distributors , and food retailers [4] , [5] . In case of a foodborne disease outbreak , rapid identification of contaminated products is essential , since the medical and economic damages incurred grow with the duration of the outbreak . Currently public health investigators must reconstruct the relevant food distribution network in order to identify the contaminated food product or contaminated product groups during an outbreak [6] . Lab-based analytical methods frequently provide the “gold standard” in verifying the source of foodborne illness outbreaks . These methods verify or cast doubt on epidemiological findings originating from case control studies with food consumption questionnaires [7] . In addition , the ability to track food through different stages of production , processing , and distribution ( traceability ) has been the subject of extensive study [8] , [9] . Nevertheless the time required to accomplish such investigations usually ranges from weeks to months . Accelerating this process may reduce the number of people sickened and help to restore consumer confidence in the safety of food products [10] . In a previous study , as a possible strategy to achieve this goal , we proposed a likelihood-based method that could be applied as an early response system to help determine the product most likely to be associated with a foodborne disease outbreak [11] . The method was tested with synthetic food sales data , but real data is readily available from retail sales companies . Proactive analysis of this retail data could complement and guide laboratory testing and trace back analysis . In the work reported here , we test our likelihood-based method using raw food sales data . As a simplifying assumption , we model food consumption at the point of sale region . In future work , we will test this assumption by applying Huff's “gravity model” for retail shopping to smooth the sales distribution over other regions [12] . Smoothing the sales distribution will also allow sensitivity analysis to spatial noise in the case reports . In applying the likelihood-based method to real world sales data , we use a ROC ( receiver operating characteristics ) analysis to quantify the performance of the method , comparing two different classifiers . This analysis also identifies the optimal discrimination threshold to maximize performance as a function of both the selectivity and specificity for the likelihood-based analysis . Additionally we explore how the method's performance may depend on “structural” properties of the sales data distribution , as this understanding is essential for efforts to proactively predict which contaminated foods/food groups might be hard to pinpoint in the event of an outbreak .
We apply product specific retail sales data from stores of a German food retail company covering 3 , 513 of Germany's 8 , 235 postal zones . The dataset lists the weekly sales of 580 anonymous food products ( N = 580 ) . For application in this analysis , sales data were aggregated per postal zone and product over the three-year period 01/2008 to 12/2010 . Let sales ( n , r ) represent the number of units of food product n sold in region r over this three-year period . We can now define a function fs ( n , r ) representing the probability that product n is sold in region r as: ( 1 ) where R is the set of all regions included in the analysis . The underlying assumption of outbreak pattern generation is that for each product the distribution of sales across the postal codes reflects the true consumption pattern for that food [12] . Hence , the function fc ( n , r ) represents the probability that product n is consumed in region r and in this paper we simply assume probability of consumption equals probability of sale: ( 2 ) Notice that for a given product n , fc ( n , r ) is a discrete probability mass function representing the probability that product n is consumed in location r , and that: ( 3 ) We take advantage of this when generating synthetic outbreak case reports for a selected “contaminated” product x ( where we use x instead of n to indicated a single contaminated product ) . Using A . J . Walker's alias method [13] , we draw M random locations by sampling from fc ( x , r ) over all locations r in R . In separate trials , synthetic case report data are generated assuming each of the 580 products , in turn , as the source of contamination . We assume the products are independent so fc ( x , r ) also defines the probability of a case report at location r due to contaminated product x . It is true that two “products” with different local “brands” or “ids” could in fact be the same food item simply rebranded when repackaged locally . Conversely , a product sold on a national scale under one single brand could become contaminated at a single point of sale retail site ( e . g . , a butcher shop ) . For the purposes of this study , the simulated case reports were generated self consistently from the retail data using the assumption that the data provided to us by product id were independent . Depending upon the spatial distribution of product x , it is likely that , during one simulated outbreak of 100 cases , multiple case reports will come from a same postal code . Figure 1 plots the number of case reports per location for several different outbreaks each generated based on a different product . Distributions generated from widely distributed products ( shown in blue ) are flatter than distributions generated from products sold only locally or regionally ( shown in red ) . An outbreak can be described by the set of locations {R} of all reported cases where ri is the location of the ith case . Note that there is no limit or constraint on how many cases may come from a particular location . In order to identify implicated products we describe two estimation methods below . We run the analysis varying the contaminated product , x , over all N = 580 products , and up to M = 100 synthetic case reports ending up with 58 , 000 vectors . Next , we repeat the experiment over S = 100 randomly seeded runs , denoting the outcome of in the sth experiment . ( See Dataset S1 and Dataset S2 . ) Now we define a statistic Ax , m as such: ( 7 ) is a function that returns 1 if the index of maximum element in vector v is i; if not it returns 0 . We call statistic A the success rate [11] . Statistic B is based upon an ROC analysis . In an ROC analysis , we compute the average true positive rate and false positive rate ( also called sensitivity and specificity ) . The average true positive rate ( TPR ) for a discrimination threshold t is defined as: ( 8 ) Here we assume the ≥ test returns 1 when satisfied , 0 otherwise . Essentially we sum the total number of outcomes where the ratio of “guilty” product x is above the threshold and then average over the S runs . To define the false positive rate for a contaminated product x after m case reports , we first compute the number of true negatives in run s: ( 9 ) Next we compute the number of false positives: ( 10 ) The average false positive rate is now defined as: ( 11 ) In the analysis , we use the thresholds t of 1/256 , 1/128 , 1/64 , 1/32 , 1/16 , 1/8 , 1/4 , 1/2 and 1 to generate the Area Under Curve ( AUC ) statistic . As some food distributions within the data set had no overlap with the generated outbreak pattern , and to avoid overestimation of specificity , we exclude so-called “zero probability” products from the average in the corresponding scenario . A product belonged to the zero probability category , by definition , when after 100 trials and 100 case reports for each trail , that product is never sold in any sampled location . Failure to exclude the zero probability set would artificially exaggerate the specificity of the method . In order to analyze how different food distribution patterns can influence the performance of the likelihood-based method , the similarity of the distribution patterns of the food products was measured by calculating the pair-wise Spearman's rank correlation coefficient , , on the basis of sales distribution data of all food products [14] . Similar to the estimation technique described in method 2 above , let sales ( k ) be a vector of length 3 , 513 ( number of locations ) where each element is the total sales of product k in a given location . The pair-wise Spearman's rank correlation is between two products , k and l becomes: ( 12 ) Since Spearman's provides a measure of pair-wised association between food distributions , the value of 1− served as a dissimilarity measure describing the “distance” between each pair of food products . This measure was used as input for a hierarchical clustering algorithm [function hclust ( ) ] using the complete linkage method provided by the R-Manual [15] .
In order to evaluate its performance the method has been applied to a real world dataset of 580 food products with known distribution patterns across Germany [11] . In this analysis the simplifying assumption has been made , that exactly one of the known food products is responsible for a disease outbreak , which were generated based on the corresponding “guilty” food product distribution . The number of sampled cases defining the outbreak size has been varied from 1 to 100 . To assess the performance of the likelihood-based method statistic A and B were used . Each statistic describes the capability of the method to correctly identify the source of infection back from the comparison of the artificial outbreak pattern with each of the 580 food products under investigation . In Figure 2 the green curve shows the success rate ( statistic A ) averaged over the 580 food products as contamination source . The average success rate of the algorithm rises steeply with the number of case reports reaching a level above 80% with only 50 case reports . However there are outbreak patterns for which the likelihood method is not effective with many more case reports required for unique identification of the correct “guilty” product . This is in line with the expectation the highest likelihood criteria are hard to accomplish for similarly distributed products . Taking advantage of the likelihood-based approach we can also assess the relative probability for all products . Selecting a discrimination threshold , we can then identify the group or subset of all products with likelihood ratio greater than that threshold , which we call the “suspect product set” . In Figure 2 we also show the average probability that the contaminated product is found within this set for thresholds of 1/8 ( cyan ) and 1/32 ( red ) . Also shown in Figure 2 is the number of products found ( on average ) within the suspect product set , as a function of the number of case reports , for the same choices of threshold . Even for a threshold of 1/32 , the average set size falls to as few as a dozen suspect products within only ten case reports [16] . To visualize the performance statistic B of this likelihood-based approach , we plot in Figure 3a the “receiver operating characteristic” or ROC curves for outbreak patterns with different numbers of cases . The ROC analysis characterizes the performance of the algorithm when the calculated likelihood ratio is applied as a binary classifier . The curve shows the “sensitivity” of the classifier as a measure of the fraction of true positives vs . the fraction of false positives ( 1-specificity ) . An ideal or perfect classifier would have a sensitivity of 1 . 0 at ( 1-specifity ) = 0 ( no false positives ) . The area under the ROC curve ( AUC ) provides a measure of overall performance . A perfect classifier has an AUC = 1 . 0 . A useless classifier ( e . g . , with a linear ROC curve and slope of ½ ) would have an AUC of 0 . 5 . As expected , this type of performance measure illustrates that the results of the likelihood-based approach depend on the number of case reports . Thus separate curves are shown for outbreaks with 1 , 2 , 3 , 5 , 10 , and 50 cases . ( The ROC curve is defined for only one case report . However , from a public health perspective an “outbreak” of foodborne illness is declared only after two or more cases . ) As Figure 3a shows , the area under the cure approaches 1 for outbreak patterns with as few as 50 case reports . In Figures 3b and 3c , we compare the performance of the likelihood-based approach with a simple classifier based on the Spearman rank correlation coefficient . As these three figures make evident , the likelihood-based method outperforms the correlation-based approach . In a real world application , these performance improvements are of utmost importance to avoid false accusations of food manufacturers , unjustified product recalls , and a waste of limited analytical resources . Using the Spearman's rank correlation coefficient , we explore how the performance of the likelihood-based method is related to associations between distinct product sales distributions . As Figure 4 and Figures S1 and S2 confirm , the algorithm's performance is strongly influenced by associations between food sales distributions . Plotting the maximum Spearman's for each product against success rate , we assess how the magnitude of the association between the contaminated product and the food to which it is most similarly distributed affects the suspect product set size determined by the likelihood-based approach . The data in Figure 4 demonstrates that the number of suspect products increases steeply if the contaminated food and the product most related to it have high correlation . The knee of the curve shifts with set size increasing sharply for correlation ≳ 0 . 8 given 10 case reports , ≳ 0 . 9 given 20 case reports , etc . Comparing Figure 4c to 4a , it is clear that as the maximum pairwise correlation between a contaminated product and another product increases , the number of cases required to reduce the suspect product set size to a manageable number ( e . g . , below 10 ) increases . In Figure S2 we also show the corresponding decrease in the “success rate” measure . Consider ‘Y’ products with identical sales distributions . When the rank ordered distribution patterns of the contaminated food and at least one of these foods are equal , then the value of the likelihood for those products will remain the same . In this limit , the size of the suspect product set will never fall below Y , independent of the number of case reports . Understanding the maximum number of highly correlated products is therefore important given the larger goal of accelerating foodborne disease investigations , as it forewarns public health investigators of the largest number of products that may have to be tested together ( in a worst case scenario ) . As noted , a high degree of similarity between the distribution patterns of the food products under investigation and the spatial pattern of the contaminated “guilty” product implies that it is ( will be ) difficult to correctly identify the causative food item . To describe and visualize this property of the food data set , we calculate the correlation matrix and apply hierarchical clustering algorithms . Figure 5 is a graphical representation of the pair-wise Spearman's correlation coefficient matrix as a so-called heat map . In this representation , products were sorted by the hierarchical clustering indicated in Figure 6 . The colors indicate the degree of similarity between food products as measured by the Spearman's . This representation supports the finding that there is a large cluster of highly similar distributed food products within the given data set . Products belonging to this cluster make the biggest contribution to rapid decrease in classifier performance when the number of case reports falls below 10 ( data not shown ) . The figure shows a distribution of cluster sizes within the retail sales data . Figure 6 shows a dendrogram visualizing the dissimilarity of the spatial distribution patterns of the 580 food products under investigation . Most similarly distributed food products are grouped at the bottom of the tree with a dissimilarity score close to 0 ( i . e . , the spatial distribution pattern is almost identical ) . Clusters of similar distributed food products are connected according to the dissimilarity score generated by the complete linkage method . For further investigations distinct clusters were generated ( indicated by different colors ) by cutting the tree horizontally at the dissimilarity level of 0 . 25 . This ensures that within each cluster , all pair-wise Spearmen's correlation coefficients are at least 0 . 75 . This choice of threshold was inspired by the observation reported above , which the suspect product set size increases rapidly when the maximum Spearman's is above ∼0 . 8 . The product data used in this study was provided as point of sale retail data by anonymized product id . After completion of the study , the products were identified as various dairy products . The 580 food items include some items that are locally branded ( and sold ) and some very widely distributed products sold nationally . The only factor we could identify as important to the product clustering shown in Figure 6 was the spatial pattern of the food distribution including whether the food item was sold locally , regionally , or nationally . Categories , such as fresh or frozen , do not affect the observed clustering ( and those factors where not used in generating the simulated outbreaks as they were not known to the authors before the study and not built into the simulation ) . To characterize the clusters observed in the dendrogram in Figure 6 , we show in Figure 7 , for all clusters containing three food products , a series of small images showing color coded product sales volume in each of the 3 , 513 postal code regions where food is sold by the food retail company . The images are organized according to the product grouping generated by the clustering algorithm . The figure clearly shows that product clustering strongly depends on how widely spread or how localized is the spatial sale pattern of the product for each cluster . Products with similar sales distributions are placed in common clusters by the pair-wise Spearman's rank correlation method . Figures 8 a–c show that the average success rate for identification of contaminated foods within a cluster of a certain size is linearly related to the log of cluster size for ( a ) 10 case reports , ( b ) 20 case reports , and ( c ) 50 case reports . It can be stated , that the absolute magnitude of the slope of this linear relationship decreases in the presence of larger numbers of cases . This confirms , that even for highly correlated food distribution patterns the performance of a likelihood-based classifier will increase with additional information on case reports . This analysis shows how , when information on the food distribution channels is available , likelihood-based methods can quickly identify those products likely to be causing an outbreak using the geographic locations for even relatively few cases . However , these methods assume that food distribution channels are well characterized , which may rarely be the case . Nevertheless , our methods could be extremely useful for retail companies that want to assess which of their own products could potentially be involved in an ongoing disease outbreak , or identifying chains or individual stores that should be prioritized for investigation in an ongoing outbreak . In practice , multiple products may be contaminated by a single food ingredient . Here we use a very simple model of the probability of individuals consuming food for particular shops , which may be quite different from real consumption patterns . In this paper we also make the simplifying assumption that food is consumed where it is sold . In fact , people travel . In the future , it is possible to extend the current work by adding Huff's “gravity model” for retail shopping behavior [12] . This will effectively smooth the sales distribution over nearby regions . It will also make it possible to test the addition of noise in the case report generator . In the simplified model , any case report occurring in region where a product is never sold ( probability 0 ) immediately excludes that product from consideration . The performance of the likelihood-based method in these more challenging scenarios will be explored in future research . This analysis also provided some fundamental insights into the relationship of method's performance and inherited properties of the analyzed food sales data . We could confirm that the degree in similarity of the spatial food distribution pattern determines how quickly the likelihood method will converge on a finite suspect product set size . Generally , the maximum pair-wise correlation with the actual contaminated product is negatively related to success rate , and positively related to the number of cases required for a perfect prediction . This suggests that it may be beneficial to consider identifying groups of products as likely to contain the tainted food , rather than focusing on finding one product . Additionally it has been shown that relevant intrinsic properties of the food sales data can be visualized by performing hierarchical clustering algorithms . This method provides a helpful graphical summary of the spatial similarity of food distributions . Further , on the basis of clusters generated by this algorithm , it is shown that log cluster size has a negative , linear relationship with success rate . This suggests that , as the number of products similarly distributed as the contaminated product increases , our ability to consistently identify the contaminated food in a small number of cases decreases . Highly correlated food product distributions are associated with products that are ( and will be ) harder to identify than uncorrelated product distributions . Since correlated product clusters can be identified proactively , suspect products can also be grouped for analysis accelerating an outbreak investigation . | Response to foodborne disease outbreaks is complicated by globalization of our food supply chains . Rapid identification of contaminated products is essential to limit the damage caused by foodborne disease . Worldwide , foodborne disease outbreaks are responsible for $9B a year in medical costs and over $75B in economic losses . Yet relevant data required to accelerate the identification of suspicious food already exists as part of the inventory control systems used by retailers and distributors today . Combining this retail data with public health case reports has the potential to hasten outbreak investigations and provide public health investigators with better information on suspected products to test . This paper demonstrates the feasibility of the principle and efficiency of this approach . Based on these findings it can be concluded that in foodborne disease outbreaks retail data could be used to speed and target public health investigations and consequently reduce numbers of sick/dead people as well as reduce economic losses to the industry . | [
"Abstract",
"Introduction",
"Methods",
"Results/Discussion"
]
| [
"algorithms",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"computer",
"and",
"information",
"sciences",
"medicine",
"and",
"health",
"sciences",
"environmental",
"health",
"mathematics",
"epidemiology",
"health",
"care",
"applied",
"mathematics",
"biology",
"and",
"life",
"sciences",
"infectious",
"disease",
"control",
"physical",
"sciences",
"computerized",
"simulations",
"agricultural",
"production",
"agriculture"
]
| 2014 | A Likelihood-Based Approach to Identifying Contaminated Food Products Using Sales Data: Performance and Challenges |
Previous studies of the genetic landscape of Ireland have suggested homogeneity , with population substructure undetectable using single-marker methods . Here we have harnessed the haplotype-based method fineSTRUCTURE in an Irish genome-wide SNP dataset , identifying 23 discrete genetic clusters which segregate with geographical provenance . Cluster diversity is pronounced in the west of Ireland but reduced in the east where older structure has been eroded by historical migrations . Accordingly , when populations from the neighbouring island of Britain are included , a west-east cline of Celtic-British ancestry is revealed along with a particularly striking correlation between haplotypes and geography across both islands . A strong relationship is revealed between subsets of Northern Irish and Scottish populations , where discordant genetic and geographic affinities reflect major migrations in recent centuries . Additionally , Irish genetic proximity of all Scottish samples likely reflects older strata of communication across the narrowest inter-island crossing . Using GLOBETROTTER we detected Irish admixture signals from Britain and Europe and estimated dates for events consistent with the historical migrations of the Norse-Vikings , the Anglo-Normans and the British Plantations . The influence of the former is greater than previously estimated from Y chromosome haplotypes . In all , we paint a new picture of the genetic landscape of Ireland , revealing structure which should be considered in the design of studies examining rare genetic variation and its association with traits .
Situated at the northwestern edge of Europe , Ireland is the continent’s third largest island , with a modern-day population of approximately 6 . 4 million . The island is politically partitioned into the Republic of Ireland and Northern Ireland , with the latter forming part of the United Kingdom ( UK ) alongside the neighbouring island of Britain . Alternative divisions separate Ireland into four provinces reflecting early historical divisions: Ulster to the north , including Northern Ireland; Leinster ( east ) ; Munster ( south ) and Connacht ( west ) . Humans have continuously inhabited Ireland for around 10 , 000 years [1] , though it is not until after the demographic upheavals of the Early Bronze Age ( circa 2200 BCE ) , that strong genetic continuity between ancient and modern Irish populations is observed [2] . Linguistically , the island’s earliest attested language forms part of the Insular Celtic family , specifically the Gaelic branch , whose historic range also extended to include many regions of Scotland , via maritime connections with Ulster [3 , 4] . A second branch of Insular Celtic , the Brittonic languages , had been spoken across much of Britain up until the introduction of Anglo-Saxon in the 5th and 6th centuries , by which time they were diversifying into Cornish , Welsh and Cumbric dialects [5] . Since the establishment of written history , numerous settlements and invasions of Ireland from the neighbouring island of Britain and continental Europe have been recorded . This includes Norse-Vikings ( 9th-12th century ) , especially in east Leinster , and Anglo-Normans ( 12th-14th century ) , who invaded through Wexford in the southeast and established English rule mainly from an area later called the Pale in northeast Leinster [6] . There has also been continuous movement of people from Britain , in particular during the 16-17th century Plantation periods during which Gaelic and Norman lands were systematically colonized by English and Scottish settlers . These events had a particularly enduring impact in Ulster in comparison with other planted regions such as Munster . As with the previous Norman invasion , the less fertile west of the country ( Connacht ) remained largely untouched during this period . The genetic contributions of these migratory events cannot be considered mutually independent , given that they derive from either related Germanic populations ( such as the Vikings and their purported Norman descendants ) or from other Celtic populations inhabiting Britain , which had themselves been subjected to mass Germanic influx from Anglo-Saxon migrations and later Viking and Norman invasions [7] . Moreover , each movement of people originated from northern Europe , a region which had witnessed a mass homogenizing of genetic variation during the migrations of the Early Bronze Age , possibly linked to Indo-European language spread . [8 , 9] . However , each event had a geographic and temporal focal point on the island , which may be detectable in local population structure . Previous genome-wide surveys have detected little to no structure in Ireland using methods such as principal component analysis ( PCA ) on independent markers , concluding that the Irish population is genetically homogenous [10] . However , runs of homozygosity are relatively long and frequent in Ireland [10] and correlate negatively with population density and diversity of grandparental origins [11] , suggesting that low ancestral mobility may have preserved regional genetic legacies within Ireland , which may be detectable in modern genomes as local population structure embedded within haplotypes . This is further supported by the restricted regional distributions of certain Y chromosome haplotypes [12 , 13] . The haplotype-based methods ChromoPainter and fineSTRUCTURE [14] were recently used to uncover hidden genetic structure among the people of modern Britain [7] . These approaches exploit the rich information available within haplotypes ( usually statistically phased ) to identify clusters of genetically distinct individuals with a resolution that could not be attained using single-marker methods . In doing so , the People of the British Isles ( PoBI ) study was able to identify discrete genetic clusters of individuals that strongly segregate with geographical regions within Britain , though notably , structure was undetectable across a large southeastern portion of the island . However , although this study sampled over 2 , 000 individuals , only 44 were from Northern Ireland with none from the remainder of the island . Ireland was also excluded from admixture and ancestry analyses due to the confounding effects of the island acting as “a source and a sink for ancestry from the UK” . With this focus on a single island , the PoBI study has an obvious limit , despite its title . Here , we have used the methods of the PoBI study to explore fine-grained Irish population substructure . We first investigate Ireland on its own , then we consider the genetic substructure observed on the island in the context of Britain and continental Europe . Using modern individuals from these two sources as surrogates for historical populations , we apply the GLOBETROTTER model to infer admixture events into Ireland and we consider these in the context of historically recorded invasions and migrations . Our inclusion of Irish data with previously-published data from Britain presents a more complete representation of genetic ancestry in the contemporary populations of the British Isles , providing a comprehensive population genetic perspective of the peopling of these islands .
We used ChromoPainter [14] to identify haplotypic similarities within a genome-wide single nucleotide polymorphism ( SNP ) dataset of individuals from the Republic of Ireland and Northern Ireland ( n = 1 , 035 , including 44 from the PoBI study ) , in which local geographic origin was known for a subset ( n = 588 ) . Clustering the resulting coancestry matrix using fineSTRUCTURE identified 23 clusters , demonstrating local population structure within Ireland to a level not previously reported , with apparent geographical , sociopolitical and ancestral correlates ( Fig 1 ) . All clusters were robustly defined , with total variation distance ( TVD ) p-values less than 0 . 001 ( S1 and S2 Tables ) . We projected the ChromoPainter coancestry matrix in lower-dimensional space using principal component analysis ( PCA ) and , to ease interpretation and for visual brevity with labels , we defined 9 cluster groups that formed higher order clades in the fineSTRUCTURE dendrogram , overlapped in PC space and were sampled from geographically contiguous regions . These cluster groups also showed robust definition by TVD analysis ( S3 Table and S4 Table ) , suggesting they represent a meaningful grouping of the data . ChromoPainter PCA revealed a tight relationship between haplotypic similarity and geographical proximity , with ChromoPainter principal component ( PC ) 1 roughly describing a north to south cline and PC2 largely describing an east to west cline ( Fig 1B ) . At a high level , both ChromoPainter PCA and fineSTRUCTURE clustering loosely separated the historical provinces of Ireland ( Ulster , Leinster , Munster and Connacht ) suggesting that these socially constructed territories may have had an impact on genetic structure within Ireland which is deeply embedded in time . Careful inspection of the tree ordering and the PCA revealed more nuanced relations between the provinces; for example south Leinster clusters share more haplotypes with those from north Munster than with their central and north Leinster counterparts . The geographical distribution of this deep subdivision of Leinster resembles pre-Norman territorial boundaries which divided Ireland into fifths ( cúige ) , with north Leinster a kingdom of its own known as Meath ( Mide ) [15] . However interpreted , the firm implication of the observed clustering is that despite its previously reported homogeneity , the modern Irish population exhibits genetic structure that is subtly but detectably affected by ancestral population structure conferred by geographical distance and , possibly , ancestral social structure . ChromoPainter PC1 demonstrated high diversity amongst clusters from the west coast , which may be attributed to longstanding residual ancient ( possibly Celtic ) structure in regions largely unaffected by historical migration . Alternatively , genetic clusters may also have diverged as a consequence of differential influence from outside populations , as this diversity between western genetic clusters cannot be explained in terms of geographic distance alone . South Munster ( SMN ) and Cork ( CRK ) clusters branch off first in the fineSTRUCTURE tree and show distinct separation from their neighbouring north Munster clusters ( NMN ) , indicating that south Munster’s haplotypic makeup is more distinct from its neighbouring regions and the remaining regions than any other cluster . TVD analysis supports this observation ( S1 Table and S3 Table ) , with the Cork cluster in particular showing strong differentiation from other clusters . This may reflect the persistent isolating effects of the mountain ranges surrounding the south Munster counties of Cork and Kerry , restricting gene flow with the rest of Ireland and preserving older structure . In contrast to the west of Ireland , eastern individuals exhibited relative homogeneity; a similar pattern was observed in the PoBI study [7] , in which all samples in a large region in southeast England formed a single indivisible cluster of genetically similar individuals comprising almost half the dataset . However , while east coast clusters in Ireland are the largest and demonstrate strong cluster integrity , the largest of these ( Central Leinster , CLN ) comprises roughly a fifth of our dataset ( S1 Fig ) , hence they are dwarfed proportionally in both number and geographical extent by the southeast England cluster ( SEE ) , suggesting that deeper structure persists in eastern Ireland than in southeast England . The overall pattern of western diversity and eastern homogeneity in Ireland may be explained by increased gene flow and migration into and across the east coast of Ireland from geographically proximal regions , the closest of which is the neighbouring island of Britain . To explore this , we estimated the extent of admixture per individual in the Irish dataset from Britain , using samples from the PoBI dataset as a reference [7] , along with eighteen ancient British individuals from the Iron Age , Roman and Anglo-Saxon periods in northeast and southeast England [16 , 17] . Using an unsupervised ADMIXTURE analysis [18] , we observed that one of the ADMIXTURE clusters ( k = 2 ) comprises the totality of ancestry of several Anglo-Saxon individuals and forms the largest proportion in British groups , with varying representation across Irish clusters ( S8 Fig ) . For simplicity we will call this the British component , which was among the lowest for individuals falling in Irish west coast fineSTRUCTURE clusters , including the south Munster and Cork cluster groups ( Fig 1D ) , supporting the interpretation that these regions differ in terms of restricted haplotypic contribution from Britain . Analysis of variance of the British admixture component in cluster groups showed a significant difference ( p < 2×10−16 ) , indicating a role for British Anglo-Saxon admixture in distinguishing clusters , and ChromoPainter PC2 was correlated with the British component ( p < 2×10−16 ) , explaining approximately 43% of the variance . PC2 therefore captures an east to west Anglo-Celtic cline in Irish ancestry . This may explain the relative eastern homogeneity observed in Ireland , which could be a result of the greater English influence in Leinster and the Pale during the period of British rule in Ireland following the Norman invasion , or simply geographic proximity of the Irish east coast to Britain . Notably , the Ulster cluster group harboured an exceptionally large proportion of the British component ( Fig 1D and 1E ) , undoubtedly reflecting the strong influence of the Ulster Plantations in the 17th century and its residual effect on the ethnically British population that has remained . The genetic substructure observed in Ireland is consistent with long term geographic diversification of Celtic populations and the continuity shown between modern and Early Bronze Age Irish people [2] . However , this diversity is weaker on the east coast in a manner that correlates with British admixture , suggesting a role for recent migrations in eroding this structure . We therefore further investigated the relationship between Ireland and Britain by generating a ChromoPainter coancestry matrix for all Irish and PoBI data combined ( n = 3 , 008 ) . Clustering with fineSTRUCTURE revealed 50 distinct clusters that segregated geographically , both on a cohort-wide and local level ( Fig 2 ) . Projecting this coancestry matrix in PC space revealed a striking concordance between haplotypes and geography ( sampling regions were defined using Nomenclature of Territorial Units for Statistics 2010; [19] ) for ChromoPainter PCs 1 and 4 , reminiscent of previous observations for single marker-based summaries of genetic variation within European populations [20] . The principal split in the combined Irish and British data defined two genetic islands , both in the fineSTRUCTURE tree and in ChromoPainter PC1 ( Fig 2 ) . This distinction between Irish and British genetic data was particularly pronounced when we applied t-distributed stochastic neighbour embedding ( t-SNE ) [21] to the ChromoPainter coancestry matrix ( Fig 3 ) . t-SNE is a nonlinear dimensionality reduction method that attempts to provide an optimal low-dimensional embedding of data by preserving both local and global structure , placing similar points close to each other and dissimilar points far apart . In principle , a two-dimensional t-SNE plot can therefore summarize more of the overall differences between groups than those described by any two principal components , although the relative group sizes , positions and distances on the plot are less straightforward to interpret . Applying t-SNE to the Irish and British coancestry matrix captured the salient structure described by ChromoPainter PCA , and particularly validates that observed in the plot of PC1 vs PC4 . This clearly distinguishes the two islands , discerns their north-south and west-east genetic structure and places Orkney and north/south Wales , whose variation is captured in PCs 2 and 3 respectively ( Fig 4 ) , as independent entities from the bulk of the British data . As observed in Fig 1 , ChromoPainter PCA in Ireland and Britain ( Fig 2 ) demonstrates eastern homogeneity for each island and relative diversity on the west coast . The southeast England ( SEE ) cluster group is centred at zero on PC4 , representing a group with predominantly Anglo-Saxon-like ancestry ( S8 Fig ) . Clusters representing Celtic populations harbouring less Anglo-Saxon influence separate out above and below SEE on PC4 . Notably , northern Irish clusters ( NLU ) , Scottish ( NISC , SSC and NSC ) , Cumbria ( CUM ) and North Wales ( NWA ) all separate out at a mutually similar level , representing northern Celtic populations . The southern Celtic populations Cornwall ( COR ) , south Wales ( SWA ) and south Munster ( SMN ) also separate out on similar levels , indicating some shared haplotypic variation between geographically proximate Celtic populations across both Islands . It is notable that after the split of the ancestrally divergent Orkney , successive PCs describe diversity in British populations where “Anglo-saxonization” was repelled [22] . PC3 is dominated by Welsh variation , while PC4 in turn splits North and South Wales significantly , placing south Wales adjacent to Cornwall and north Wales at the other extreme with Cumbria , all enclaves where Brittonic languages persisted . Scotland is another region of Britain which successfully retained its Celtic language , however in contrast to Welsh and Cornish clusters , the majority of Scottish variation is described by ChromoPainter PC1 . The three definable Scottish groups do not drive any further components of variation ( up to PC7 considered ) and fall away from the bulk of British variation on PC1 towards Irish clusters . This is most strikingly observed for the southern Scottish cluster ( SSC ) which fell amongst Irish branches in the fineSTRUCTURE tree , overlapping with samples from the north of Ireland in PC space ( Fig 2 and Fig 5 ) . In an interesting symmetry , many Northern Irish samples clustered strongly with southern Scottish and northern English samples , defining the Northern Irish/Cumbrian/Scottish ( NICS ) cluster group . More generally , by modelling Irish genomes as a linear mixture of haplotypes from British clusters , we found that Scottish and northern English samples donated more haplotypes to clusters in the north of Ireland than to the south , reflecting an overall correlation between Scottish/north English contribution and PC1 position in Fig 1 ( Linear regression: p < 2×10−16 , r2 = 0 . 24 ) . North to south variation in Ireland and Britain are therefore not independent , reflecting major gene flow between the north of Ireland and Scotland ( Fig 5 ) which resonates with three layers of historical contacts . First , the presence of individuals with strong Irish affinity among the third generation PoBI Scottish sample can be plausibly attributed to major economic migration from Ireland in the 19th and 20th centuries [6] . Second , the large proportion of Northern Irish who retain genomes indistinguishable from those sampled in Scotland accords with the major settlements ( including the Ulster Plantation ) of mainly Scottish farmers following the 16th Century Elizabethan conquest of Ireland which led to these forming the majority of the Ulster population . Third , the suspected Irish colonisation of Scotland through the Dál Riata maritime kingdom , which expanded across Ulster and the west coast of Scotland in the 6th and 7th centuries , linked to the introduction and spread of Gaelic languages [3] . Such a migratory event could work to homogenise older layers of Scottish population structure , in a similar manner as noted on the east coasts of Britain and Ireland . Earlier communications and movements across the Irish Sea are also likely , which at its narrowest point separates Ireland from Scotland by approximately 20 km . To temporally anchor the major historical admixture events into Ireland we used GLOBETROTTER [23] with modern surrogate populations represented by 4 , 514 Europeans [24] and 1 , 973 individuals from the PoBI dataset [7] , excluding individuals sampled from Northern Ireland . Of all the European populations considered , ancestral influence in Irish genomes was best represented by modern Scandinavians and northern Europeans , with a significant single-date one-source admixture event overlapping the historical period of the Norse-Viking settlements in Ireland ( p < 0 . 01; fit quality FQB > 0 . 985; Fig 6 ) . This was recapitulated to varying degrees in specific genetically- and geographically-defined groups within Ireland , with the strongest signals in south and central Leinster ( the largest recorded Viking settlement in Ireland was Dubh linn in present-day Dublin ) , followed by Connacht and north Leinster/Ulster ( S5 Fig; S6 Table ) . This suggests a contribution of historical Viking settlement to the contemporary Irish genome and contrasts with previous estimates of Viking ancestry in Ireland based on Y chromosome haplotypes , which have been very low [25] . The modern-day paucity of Norse-Viking Y chromosome haplotypes may be a consequence of drift with the small patrilineal effective population size , or could have social origins with Norse males having less influence after their military defeat and demise as an identifiable community in the 11th century , with persistence of the autosomal signal through recombination . European admixture date estimates in northwest Ulster did not overlap the Viking age but did include the Norman period and the Plantations ( S5 Fig ) . This may indicate limited Viking activity in Ulster , or , that due to the similarity in sources for the Viking and Anglo-Norman invasions and the Plantations , GLOBETROTTER failed to disentangle the earlier events from the later . This is not unexpected given the extent of the Plantations in Ulster [26] , the relative timings of the invasions and the degree of Viking involvement in Britain and Europe . Indeed , when considering Britain as an admixing source using PoBI data , GLOBETROTTER date estimates for northwest Ulster overlapped the Plantations , although for other regions in Ireland ( and for Ireland considered as a whole ) these admixture events were less clearly defined , likely reflecting a history of continuous gene flow between the two islands in the prevailing years ( Fig 6; S5 Fig and S7 Table ) . The all-Ireland point estimate for admixture from Britain spanned the Norman settlement instead of the Plantations , but GLOBETROTTER was unable to adequately resolve the model details for this event ( fit quality FQB < 0 . 985; Fig 6 ) , indicating that this estimate is not a good reflection of the true timings and extent of admixture from Britain . As noted in the PoBI study , the overall influence of British admixture in Ireland ( and vice versa ) has involved extensive and constant gene flow before , during and after the major population movements detailed in Fig 6 , with particular swells of peopling during the Plantations . The genetic legacies of the populations of Ireland and Britain are therefore extensively intertwined and , unlike admixture from northern Europe , too complex to model with GLOBETROTTER . Our results show that population structure is detectable on the island of Ireland and is consistent with a combination of the homogenizing effect of geographically punctuated admixture and diversification among Celtic subpopulations . The inclusion of Irish data with British samples from the PoBI study provides an anchor for Celtic ancestry in the British Isles , filling out the genetic landscape of the islands . It is also clear that historical migrations into Ireland have left a greater genomic footprint than previously anticipated; our consideration of autosomal data escapes the constraints of patrilineal genetics and has allowed us to detect a much greater Viking influence than previously estimated with Y chromosome data . Although the genetic imprint of the British Plantations is much harder to delineate , the inter-island exchange and clustering observed between present-day individuals from Northern Ireland and Scotland signals the enduring impact of these historical movements of people . Unlike the PoBI study , Irish data were not specifically selected for longstanding pure ancestry in each geographic region ( for example , having four grandparents in a location ) , but instead represent a repurposed medical dataset . Our data are therefore more representative of those that are typically used in population-based genome-wide surveys for trait-associated genetic variation; as these studies survey increasingly rare genetic variants in larger populations , the geospatial segregation of rare haplotypes and variants will become increasingly important , especially when environmental effects and interactions play a role [27] . Our observation that these haplotypes are intricately tied to geography in Ireland and Britain highlights the importance of considering fine-grained population structure in future studies .
All Irish subjects provided written informed consent to participate in genetic research and the study was approved by the Beaumont Hospital Research Ethics Committee in Dublin , Ireland under approval number 05/49 following guidelines laid out at www . beaumontethics . ie . Our study included three datasets of genotype data: a population-based Irish ALS case-control dataset ( n = 991 ) incorporating existing [28] and newly-genotyped samples , the People of the British Isles dataset ( EGA accession ID EGAD00010000632; n = 2 , 020 ) [7] and a pan-European dataset derived from a genome-wide association study ( GWAS ) for multiple sclerosis ( MS; EGA accession ID EGAD00000000120; n = 4 , 514 ) [24] ( S1 Text: Populations ) . All Irish subjects provided written informed consent to participate in genetic research and the study was approved by the Beaumont Hospital Research Ethics Committee in Dublin , Ireland . We applied quality control to each dataset using PLINK 1 . 9 [29] and merged data as detailed in Supplementary Methods ( S1 Text: Quality Control ) . Briefly , we excluded both infrequent and high-missingness SNPs; individuals with high missingness , excessive heterozygosity or cryptic relationships to other individuals in the data; and finally individuals who had been removed during QC carried out in the source papers . As the European dataset included patients and controls from a GWAS for MS , we additionally removed SNPs in a 15 Mb region surrounding the strongly associated HLA locus on chromosome 6 ( GRCh37 position chr6:22 , 915 , 594–37 , 945 , 593 ) , as is consistent with previous studies using the data [7 , 30] . This was to avoid haplotypic bias arising from this association . The final post-QC Irish ( n = 991 ) , British ( n = 2 , 020 ) and European datasets ( n = 4 , 514 ) contained 407 , 750 SNPs , 521 , 883 SNPs and 363 , 396 SNPs at zero missingness , respectively . The final merge of British and Irish data ( n = 3 , 008 ) and European and Irish data ( n = 5 , 506 ) contained 214 , 632 SNPs and 166 , 139 SNPs respectively at zero missingness . Further details regarding samples and QC per dataset are described in Supplementary Methods ( S1 Text: Populations and S1 Text: QC ) Geographic information was available for 544 of the 991 Irish samples in the form of home address . To preserve anonymity this was jittered in all maps containing patients ( Fig 1 and S5 Fig ) . For all British and some Northern Irish data , sample location was supplied by the authors of PoBI [7] as membership of 35 sampling regions . Finally , for European data sampling country was available [24] . Full details of treatment of samples for mapping are available in Supplementary methods ( S1 Text: Mapping . ) We phased autosomal genotypes in each dataset and merged dataset with SHAPEIT V2 [31] using the 1000 Genomes ( Phase 3 ) as a reference panel [32] . A pre-phasing step was carried out ( —check ) to remove any SNPs which did not correctly align to the 1000 genomes reference panel . Samples were then split by chromosome and phased together using default settings and the GRCh37 build genetic map to estimate linkage disequilibrium . To detect population structure we performed ChromoPainter/fineSTRUCTURE analysis [14] on each of the population datasets ( Irish , British and European ) individually , and then separately on a merge of the Irish and British datasets . In brief , we used ChromoPainter to paint each individual using all other individuals ( -a 0 0 ) using default settings with the exception that the number of “chunks” per region value was set to 50 ( -k 50 ) for all analyses including Irish and British individuals to account for the longer haplotypes observed in these datasets , in keeping with previous studies [7 , 30] . The fineSTRUCTURE algorithm was then run on the resulting coancestry matrix to determine genetic clusters based on patterns of haplotype sharing . Further details are included in the Supplementary Methods ( S1 Text: fineSTRUCTURE analysis ) . We assessed the robustness of Irish clusters by calculating total variation distance ( TVD ) as described in the PoBI study [7] . This metric compares the “copying vectors” of pair of clusters . Here we define the copying vector for a given cluster A as a vector of the average lengths of DNA donated by each cluster to individuals within cluster A under the ChromoPainter model . Hence the magnitude of differences between copying vectors of two clusters reflects the distances between those clusters in terms of their haplotypic sharing with other clusters . TVD can therefore be used to determine whether fineSTRUCTURE clusters detect significant differences in haplotype sharing , and hence ancestry . We tested whether the observed clustering performed better than chance by permuting ( 1 , 000 times ) the individuals in each of our cluster pairings into clusters of the same size , and calculating the number of permutations that exceeded our original TVD score . If 1 , 000 unique permutations were not possible , the maximum number of unique permutations was used instead . P-values were calculated based on the number of permutations greater than or equal to the original TVD statistic . All p-values for Irish clusters were less than or equal to 0 . 001 indicating robust clustering ( S1 Table and S2 Table ) . We also applied these methods to our Irish cluster groups ( Fig 1 ) and observed that these are statistically distinct ( S3 Table and S4 Table ) . To provide an additional measure of population differentiation between “cluster groups” we calculated mean FST between groups using PLINK 1 . 9 [29] which is reported in S5 Table . We used the GLOBETROTTER method [23] to infer and date admixture events from Europe and Britain into Ireland separately . GLOBETROTTER uses output from ChromoPainter to estimate the pairwise likelihood of being painted by any two surrogate populations at a variety of genetic distances to generate coancestry curves . Assuming a single admixture event , these curves are expected to follow an exponential decay rate equal to the time in generations since admixture occurred [23] . As the true admixing sources are modelled as a linear mixture of surrogate sources rather than individual sources this method has the advantage of not requiring exactly sampled source populations . For our analysis we ran GLOBETROTTER with default settings twice to detect simple admixture into the island of Ireland as a whole , as well as into individual genetic clusters from the Republic of Ireland ( S5 Fig ) . European clusters ( S4 Fig ) and British clusters ( S3 Fig ) were used as surrogate populations to represent the admixing sources in two independent analyses . Target and donor clusters for this analysis were defined using the fineSTRUCTURE maximum concordance tree method described in PoBI [7] to ensure homogeneity ( Supplementary methods S1 Text: fineSTRUCTURE analysis ) ; hence , the Irish target clusters that were used differ slightly from those in Fig 1 . Briefly , for each surrogate population separately ( Europe and Britain ) we applied ChromoPainter v2 to paint Ireland and the surrogate population using the surrogate population as donors and generated a copying matrix ( chunk lengths ) for all individuals , and also 10 painting samples for each Irish individual as recommended . GLOBETROTTER was then run for 5 mixing iterations twice , first using the null . ind:1 setting to test for any evidence of admixture and then null . ind:0 setting to infer dates and sources . We ran 100 bootstraps for admixture date and calculated the probability of a null model of no admixture as the proportion of nonsensical inferred dates ( <1 or >400 generations ) produced by the null . ind:1 model , as in the GLOBETROTTER study [23] . Confidence intervals for the date were calculated from the bootstraps for the standard model ( null . ind: 0 ) using the empirical bootstrap method . ( See S1 Text: Globetrotter analysis of Admixture Dates for further details ) . A generation time of 28 years was assumed as in previous studies of this nature [7 , 23] for conversion of all date estimates from generations to years . We assessed the ancestral make up of Ireland in terms of Europe and Britain for each Republic of Ireland cluster ( see Estimating admixture dates ) to explore variation in ancestry across Ireland . To do so we modelled each cluster’s average genome as a linear mixture of the European and British donor populations using the method described in the PoBI study [7] and implemented in GLOBETROTTER ( num . mixing . iterations: 0 ) . This approach uses the ChromoPainter chunk length output to estimate the proportion of DNA which most closely coalesces with each individual from the donor populations , correcting for noise caused by similarities between donor populations whose splits may have occurred after the coalescence event . This is achieved through a multiple linear regression of the form Yp = B1X1 + B2X2 + … +BgXg , where Yp is a vector of the averaged length ( cM ) of DNA that individuals across cluster P copy from each donor group , normalised to sum to 1 across all donor groups , and Xg is the vector describing the average proportion of DNA that individuals in donor group g copy from other donor groups including their own . The coefficients of this equation B1…Bg are thus interpreted as the “cleaned” proportions of the genome ancestral to each donor group . The equation is solved using a non-negative-least squares function such that Bg ≥ 0 and the sum of proportions across groups equals 1 . To assess uncertainty of these ancestry proportion estimates we again follow PoBI [7] and resample from the ChromoPainter chunk length output to generate Np pseudo individuals for each cluster P . We achieve this by randomly sampling each of the autosomal chromosome pairs 1–22 with replacement Np times from the pool of all autosomal chromosomes pairs 1–22 across all individuals within that cluster , and then randomly summing sets of 22 of these chromosome pairs to generate each pseudo individual . We then use these Np pseudo individuals as a bootstrap for Yp above and solve for Bg . We resampled 1 , 000 times per cluster and used the inner 95% quantiles of this sampling distribution to estimate confidence intervals for the sample . For comparison we implemented an alternative delete one chromosome jack-knife approach as in Montinaro et al . [33] , and estimated the s . e . as in ref . [34] ( S6 Fig and S7 Fig ) . We also used this linear regression model to determine per-individual ancestry proportion estimates from different British clusters across Ireland , treating each individual as a cluster to enable us to assess whether gene flow from northern Britain had a gradient across Ireland . To estimate the proportion of British admixture into Irish clusters , ADMIXTURE [18] was run on the combined PoBI and Irish datasets , alongside eighteen ancient individuals from the Iron Age , Roman and Anglo-Saxon periods of northeast and southeast England [16 , 17] . Pseudo-haploid genotypes were generated for the ancient genomes at the relevant variant sites , as is standard for low coverage data , and subsequently merged with the modern diploid dataset . Data were then pruned for linkage disequilibrium between SNPs using PLINK 1 . 9 ( r2 > 0 . 25 in a sliding window of 1000 SNPs advancing 50 SNPs each time ) resulting in 86 , 481 remaining SNPs . No missingness was allowed for modern individuals , with a range of 33 , 643–85 , 553 sites used for ancient samples . Following ADMIXTURE estimation , cross-validation error was calculated using the—cv flag for 5 iterations to determine the K value for which the model has the best predictive accuracy ( K = 2 ) . Additionally 200 bootstraps of the data were run to estimate the standard error of the parameters using the–B flag . This British admixture component was regressed against PC2 of the Irish ChromoPainter coancestry matrix to determine the role of British ancestry in the differentiation of PC2 in Ireland . We also performed analysis of variance ( ANOVA ) on British admixture component per cluster group to identify if cluster by cluster differences existed . ChromoPainter coancestry matrices were projected in lower-dimensional space using principal component analysis ( PCA ) and t-distributed stochastic neighbour embedding ( t-SNE ) [21] . PCA was run using the default approach provided as part of the fineSTRUCTURE R tools [14] ( http://www . paintmychromosomes . com ) . The R package Rtsne ( https://github . com/jkrijthe/Rtsne ) was used to construct a 2-dimensional embedding of the ChromoPainter coancestry matrix over 5 , 000 iterations using a perplexity of 30 , a learning rate of 200 and an initial PCA calculated over 100 dimensions . Several t-SNE runs were performed to assess concordance between embedding solutions . All linear regressions and ANOVA tests were carried out in base statistics package in R version 3 . 2 . 3 [35] . | A recent genetic study of the UK ( People of the British Isles; PoBI ) expanded our understanding of population history of the islands , using newly-developed , powerful techniques that harness the rich information embedded in chunks of genetic code called haplotypes . These methods revealed subtle regional diversity across the UK , and , using genetic data alone , timed key migration events into southeast England and Orkney . We have extended these methods to Ireland , identifying regional differences in genetics across the island that adhere to geography at a resolution not previously reported . Our study reveals relative western diversity and eastern homogeneity in Ireland owing to a history of settlement concentrated on the east coast and longstanding Celtic diversity in the west . We show that Irish Celtic diversity enriches the findings of PoBI; haplotypes mirror geography across Britain and Ireland , with relic Celtic populations contributing greatly to haplotypic diversity . Finally , we used genetic information to date migrations into Ireland from Europe and Britain consistent with historical records of Viking and Norman invasions , demonstrating the signatures of these migrations the on modern Irish genome . Our findings demonstrate that genetic structure exists in even small isolated populations , which has important implications for population-based genetic association studies . | [
"Abstract",
"Introduction",
"Results",
"and",
"discussion",
"Methods"
]
| [
"geomorphology",
"biogeography",
"irish",
"people",
"ecology",
"and",
"environmental",
"sciences",
"european",
"union",
"landforms",
"population",
"genetics",
"geographical",
"locations",
"topography",
"ireland",
"ethnicities",
"multivariate",
"analysis",
"mathematics",
"statistics",
"(mathematics)",
"population",
"biology",
"united",
"kingdom",
"islands",
"research",
"and",
"analysis",
"methods",
"europe",
"geography",
"mathematical",
"and",
"statistical",
"techniques",
"principal",
"component",
"analysis",
"european",
"people",
"phylogeography",
"people",
"and",
"places",
"earth",
"sciences",
"genetics",
"population",
"groupings",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"evolutionary",
"biology",
"statistical",
"methods"
]
| 2018 | Insular Celtic population structure and genomic footprints of migration |
Chromosomal inversion polymorphisms are thought to play a role in adaptive divergence , but the genes conferring adaptive benefits remain elusive . Here we study 2La , a common polymorphic inversion in the African malaria vector Anopheles gambiae . The frequency of 2La varies clinally and seasonally in a pattern suggesting response to selection for aridity tolerance . By hybridizing genomic DNA from individual mosquitoes to oligonucleotide microarrays , we obtained a complete map of differentiation across the A . gambiae genome . Comparing mosquitoes homozygous for the 2La gene arrangement or its alternative ( 2L+a ) , divergence was highest at loci within the rearranged region . In the 22 Mb included within alternative arrangements , two ∼1 . 5 Mb regions near but not adjacent to the breakpoints were identified as being significantly diverged , a conclusion validated by targeted sequencing . The persistent association of both regions with the 2La arrangement is highly unlikely given known recombination rates across the inversion in 2La heterozygotes , thus implicating selection on genes underlying these regions as factors responsible for the maintenance of 2La . Polymorphism and divergence data are consistent with a model in which the inversion is maintained by migration-selection balance between multiple alleles inside these regions , but further experiments will be needed to fully distinguish between the epistasis ( coadaptation ) and local adaptation models for the maintenance of 2La .
Dobzhansky's studies of chromosomal inversion polymorphisms in natural populations of Drosophila provided the first evidence that selection played an indispensable role in their maintenance , helping to spark the neo-Darwinian synthesis [1 , 2] . More recent studies implicate selection in maintaining inversion polymorphisms in a diversity of eukaryotes , including humans [3–6] . A mechanism thought to facilitate their maintenance is reduced recombination . In inversion heterozygotes ( heterokaryotypes ) , recombination between alternate arrangements may be inhibited both by asynapsis and because single crossovers within an inversion loop result in aneuploid meiotic products [7] . Such reduced recombination binds together favorably interacting genes ( coadapted gene complexes ) and/or multiple genes that are individually adapted to local conditions , and stabilizes them against gene exchange with migrants from other genetic backgrounds [1 , 8] . Stabilization of these allelic combinations allows the inversion to establish and spread , and consequently organisms can become adapted to highly divergent environmental conditions . Although selection has been invoked repeatedly to explain the maintenance of chromosomal inversions—and in some cases associated phenotypic traits have been identified [4]—the genes or regions involved remain elusive . If a small subset of genes within an inversion were under selection and there was no gene flux at all between arrangements ( sensu [9] ) , it would be impossible to identify specific genes or even regions affected by selection . Fortunately , inhibition of gene exchange between alternative gene arrangements is not absolute . Except near inversion breakpoints , gene conversion is unaffected and double crossovers can result in balanced recombinant gametes [10 , 11] . Working together , both recombinational processes ( gene flux , [10] ) gradually break up linkage disequilibrium within arrangements and homogenize sequence variation between them , unless countered by selection . The interaction of gene flux and selection is expected to produce a mosaic of more- and less-differentiated regions inside the inversion and away from breakpoints , exactly the pattern observed from some molecular studies of inversion polymorphisms in natural populations ( e . g . , [12 , 13] ) . These observations suggest that regions affected by selection can be identified , if not the precise genes and mutations involved . It is expected that such regions will be in significant linkage disequilibrium with the inversion and with each other even when they are not adjacent . However , a neutral explanation for patterns of linkage disequilibrium also exists: regions significantly associated with the inversion polymorphism may simply represent historical remnants of the genetic background upon which mutations arose . These alternative hypotheses can be tested using estimates of the rate of genetic exchange in heterokaryotypes and the age of inversion polymorphism . Mapping of divergent regions between chromosomal arrangements is a prerequisite to identifying candidate genes under selection and ultimately elucidating the molecular basis of adaptations conferred by inversions . The reduced level of recombination in heterokaryotypes renders a traditional QTL ( quantitative trait locus ) mapping approach impractical . Instead , genomic scans of nucleotide divergence in natural populations take advantage of recombination over many generations . Previous scans for divergence and linkage disequilibrium in inversions have been hindered by the low resolution afforded by limited numbers of genetic markers . Application of modern genomics tools to the classical study of inversions has the potential to both accelerate and refine the mapping of diverged regions . Recent studies have demonstrated the utility of Affymetrix GeneChip arrays as high density genetic markers [14–17] . Emitted fluorescence from individual probes on the array directly correlates with the sequence similarity between hybridized DNA and the probe . In this manner , divergence between two genetic classes ( e . g . , alternative chromosomal orientations ) can be examined at high resolution . Using this technique , we examined genic differentiation between individual Anopheles gambiae mosquitoes bearing alternate chromosomal arrangements—2La and 2L+a—on the left arm of the second chromosome , as a first step in identifying candidate regions maintaining inversion 2La in natural populations . The 2La inversion system in A . gambiae is of interest not only as a model for understanding the adaptive role of inversions , but also for its epidemiological importance . A . gambiae is the most proficient vector of human malaria in the world , causing more than one million malaria-related deaths in sub-Saharan Africa each year [18] . Abundant inversion polymorphisms on chromosome 2 appear to play a key role in the ecological success of this species , as different inversion combinations are nonrandomly associated with both natural and anthropogenic environmental heterogeneities [3 , 19 , 20] . Inversion 2La is of particular interest and significance . First , it is the only inversion polymorphic on 2L , simplifying its analysis . Second , this inversion was acquired from an arid-adapted sibling species , A . arabiensis , by introgressive hybridization [3 , 21 , 22] . The nonrandom association of 2La with degree of aridity points to the adaptive value of this polymorphism in A . gambiae . In many different locations across Africa , 2La frequency exhibits strong and stable geographic clines from near fixation in arid zones to complete absence in humid rainforests [23–26] . Similarly , its frequency varies seasonally and microspatially according to patterns of rainfall and microclimate . Thus mosquito carriers of 2La are more likely than carriers of 2L+a to rest inside houses at night where a saturation deficit exists , affecting the probability of vector–human contact at peak blood feeding times ( reviewed in [27] ) . From the standpoint of epidemiology and human health , the 2La polymorphism has increased malaria transmission by A . gambiae across diverse ecoclimatic zones and it could mitigate the efficacy of control measures that assume uniform indoor resting and biting behavior , such as bednets and indoor insecticide ( or fungicide ) application . Its study is facilitated by several recent developments , including a completely sequenced reference genome [28] and a resulting Affymetrix GeneChip array already proven to be an effective population genomics tool [16] . Furthermore , the breakpoints of the 2La inversion have been characterized molecularly [29] and the rate of genetic exchange on 2L between 2La/+a heterozygotes has been estimated in laboratory crosses [30] . In the following we address the selective maintenance of the 2La inversion polymorphism by multiple experiments . ( 1 ) We applied the Affymetrix GeneChip to map , at unprecedented resolution , highly diverged regions between alternate arrangements of a chromosomal inversion ( 2La ) . ( 2 ) We validated the principal microarray findings by targeted DNA sequencing , and used the resulting nucleotide polymorphism data to ask whether the introgressed chromosomal arrangement rose to its current frequency via adaptive natural selection . ( 3 ) We used known recombination rates in inversion heterokaryotypes to assess the likelihood that linkage disequilibrium between diverged regions and the inversion is maintained by selection .
The Anopheles probes on the Affymetrix GeneChip Anopheles/Plasmodium Array were designed mainly from the reference A . gambiae genome assembled from the chromosomally standard ( uninverted ) PEST strain [28] . The 25 bp probes ( 11 per transcript ) interrogate ∼14 , 900 putative transcripts predicted from an early gene build ( GeneBuild 2 , 2003 ) . From this core set of probes , those with more than one perfect match or with single nucleotide mismatches elsewhere in the genome were excluded from the analysis . The remaining 151 , 213 unique probes were distributed across the genome roughly in proportion to chromosome arm length . The 49 Mb chromosome 2L was represented by 33 , 892 probes , of which 13 , 984 mapped within the 22 Mb 2La inversion . To minimize the contribution of ecological or geographic diversity to genetic variation , samples of A . gambiae homozygous for alternative 2L arrangements ( 2La/a and 2L+a/+a ) were collected simultaneously from one village in central Cameroon where 2La is highly polymorphic and inversion heterozygotes are common ( the 2La frequency was 46% in our 2005 sample of 70 mosquitoes , and in samples collected from the same village in 2002–2003 its frequency was 39%; F . Simard , unpublished data ) . The specimens used in this study were all identified as the S molecular form , one of two assortatively mating incipient species of A . gambiae [31] . Three of five 2La/a specimens were polymorphic for inversions on 2R; all other specimens carried the 2R standard arrangements . Labeled genomic DNA from each of five 2La/a and 2L+a/+a mosquitoes was hybridized to individual arrays ( ten in total ) to map nucleotide divergence , measured in terms of single feature polymorphisms ( SFPs ) . SFPs were defined as probes whose hybridization intensities were significantly different between the five carriers of each 2L arrangement , as determined by two-tailed t-tests with a threshold of p < 0 . 01 [16] . A significant difference in hybridization intensity between samples reflects underlying differences in the target nucleotide sequences interrogated by the probes on the array . Genome-wide , 1 , 352 probes ( 0 . 89% ) were SFPs between 2La- and 2L+a-carriers . Of these , 444 were found on 2L , 283 of which were found in the rearranged region . The proportion of SFPs was notably higher on 2L ( 1 . 29% ) than across the other four chromosome arms ( 0 . 79%; p < 1×10−21 ) , which show consistent levels of differentiation ( Figure 1 ) . Along 2L , SFPs were distributed disproportionately within as compared to outside the rearranged region ( 1 . 98% versus 0 . 80%; p < 1×10−20; Figure 2 ) . Indeed , the level of differentiation on 2L outside the rearranged region was indistinguishable from that of the other four chromosome arms ( p = 0 . 75 ) . The distribution of putative SFPs on 2L was also explored by implementing a hidden Markov model ( HMM ) to identify differentiated and homogenized regions independent of a priori information about 2La breakpoint locations ( cf . [16] ) . The HMM identified a 22 Mb diverged region corresponding almost precisely to the rearranged region , beginning at the first probe inside the proximal breakpoint and ending only ∼386 kb ( 243 probes ) beyond the distal breakpoint . To test for regional clustering of SFPs between the breakpoints of the rearrangement , a sliding window analysis was performed that revealed two regions in which the observed number of SFPs was greater than expected by chance ( Figure 3 ) . The first ( proximal ) cluster ( p = 0 . 001 ) extends ∼1 . 0 Mb , approximately from 2L coordinates 21 . 1–22 . 1 Mb and as measured from its midpoint , ∼1 . 1 Mb from the proximal breakpoint ( labeled “1” in Figure 3 ) . Though the boundaries of the clusters are necessarily imprecise , the first cluster spans roughly 32 genes and contains 17 SFPs . The second ( distal ) cluster ( p < 1×10−5 ) extends between coordinates ∼38 . 8–40 . 5 Mb , ∼2 . 5 Mb from the distal breakpoint ( labeled “2” in Figure 3 ) . The 63 SFPs in this ∼1 . 7 Mb cluster span approximately 178 genes . Both clusters remained significant ( p < 0 . 05 ) after correction for the number of windows tested . The genes predicted in each cluster according to Genebuild AgamP3 . 4 ( http://agambiae . vectorbase . org/index . php ) are listed in Supplementary Tables S2 and S3 . The microarray analysis suggested two patterns . Most striking was heightened divergence between the rearranged region and little elsewhere on 2L . In addition , two significant clusters of SFPs inside the rearrangement were found near , but not directly adjacent to , its proximal and distal ends . Because of the limited sample sizes in the microarray analysis , we sought to confirm and extend these results by targeted sequencing of an additional 24–34 chromosomes carrying each gene arrangement , sampled from the same Cameroon population of A . gambiae . Eleven genes were chosen based on their location within or outside of the rearranged region ( Figure 3; Table 1 ) . Two were located ∼1 Mb outside of the proximal and distal breakpoints; inside , one was located centrally , one within the proximal cluster , three within the distal cluster , and two just outside of and flanking each cluster . Wherever possible , the corresponding genes were also sequenced from 2–6 chromosomes of two sibling species: a sympatric population of A . arabiensis ( fixed for 2La ) and an allopatric population of A . quadriannulatus ( fixed for 2L+a ) . We used three approaches to explore whether DNA sequences supported the microarray results: comparing numbers of shared versus fixed differences between chromosomal arrangements , summary statistics of nucleotide differentiation , and gene tree reconstruction . The numbers of polymorphisms shared between alternative arrangements in A . gambiae was high at the two genes outside the rearranged region ( 24 and 31 ) , while the corresponding values for genes inside were much lower ( ranging from 1–9 ) ( Table 2 ) . A small proportion of all shared polymorphisms may be due to recurrent mutation ( Table 2; [32] ) , but most are shared because of gene flux ( see below ) . As expected , the number of fixed differences followed a trend opposite to that of shared polymorphisms . No fixed differences occurred outside the rearranged region; inside , five of nine genes had fixed differences . The three genes nearest the proximal breakpoint all show relatively high numbers of fixed differences , even a gene distal to the proximal cluster ( asph ) . These data therefore indicate that the boundaries of the proximal cluster were likely underestimated , though sequences at the breakpoint are expected to show fixed differences because of low levels of gene flux ( see below ) . At the other end of the rearranged region , two of the three genes within the distal cluster show fixed differences , while no fixed differences were observed in either of the flanking genes , not even the one nearest the distal breakpoint . The ratio of fixed differences to shared polymorphisms for three genes within the distal cluster ( cpr34 , cpr63 , srpn2; 9:14 ) was significantly higher than that for three genes outside and flanking the region ( hdac , nwk , depcd5; 0:21; Fisher's exact test , p = 0 . 001 ) . Because this test was conducted post hoc the results should be interpreted with some caution . Differentiation between arrangements was also gauged by estimating FST values and net divergence along 2L . Pairwise FST values for all loci on 2L were significantly different from zero , but values within the rearranged region were roughly an order of magnitude larger than those in collinear regions ( Table 2 ) . Net divergence ( Da ) followed the same general pattern , with far greater levels of divergence observed inside the rearranged region . The average number of net nucleotide substitutions per site was 1 . 7% between arrangements and only 0 . 1% outside them . Moreover , those genes showing the highest levels of divergence were located in and around the proximal cluster and—with the exception of cpr63—inside the distal cluster . The lowest level of divergence was measured at depcd5 nearest the distal breakpoint . Gene trees reconstructed from sequences at each locus yielded three basic patterns that were consistent with those that emerged from measures of divergence and fixed/shared variation ( Figure 4 ) . Owing to recombination , these branching patterns do not represent the exact evolutionary history of the genes sampled , but they do portray contrasting pictures of the extent of genetic exchange between arrangement classes . The first pattern , exemplified by lys-c in Figure 4A , shows the complete intertwining of sequences from inverted and standard arrangements , as expected if gene exchange has been frequent . This pattern was shared by both genes located outside of the inversion , as well as two inside at the distal end: cpr63 ( in the distal cluster ) and depcd5 . These latter two genes showed the lowest level of net nucleotide divergence of any genes inside the inversion and correspondingly reduced FST values . The second pattern , common to five genes within the inversion ( asph , endp , srpn2 , unk and cpr34 ) , shows reciprocally monophyletic 2La and 2L+a sequences , as expected if they are largely isolated . Three of four genes sampled from both clusters showed this pattern , including the srpn2 gene illustrated in Figure 4B . Two remaining genes inside the inversion ( hdac and nwk ) gave a third pattern indicative of limited gene exchange , such that a single sequence of one arrangement clustered together with sequences of the opposite arrangement as illustrated for nwk in Figure 4C ) . A notable feature of all gene trees where outgroup sequences were available was the embedding of A . arabiensis ( 2La ) and A . quadriannulatus ( 2L+a ) sequences inside of A . gambiae 2La and 2L+a clades , respectively . A . gambiae is considered to have evolved from an A . quadriannulatus-like ancestor in a recent human-influenced speciation event in the central African rain forest [33] . If so , it would have carried only 2L+a , as A . quadriannulatus does . Although the 2La arrangement is ancestral in the A . gambiae sibling species complex [29 , 33] , 2La likely passed into A . gambiae subsequent to the emergence of this species , following contact and genetic introgression with A . arabiensis [3 , 21 , 22] . Further evidence of the close genetic relationship between the same arrangement of 2L in different species can be seen from the contrasts presented in Table 2 . In the rearranged region , greater differentiation exists between alternative arrangements within A . gambiae than between the same arrangement from different species . The opposite is true for collinear regions: differentiation is greater between species than within A . gambiae , due to free recombination in the latter . Given that deeper sequencing of both chromosomal arrangements confirmed the existence of two highly differentiated clusters of genes within the rearranged part of 2L , we next asked whether there was any signature of selection in these clusters or on the 2La arrangement . The presumed recent introgression of 2La is inconsistent with a long-term balanced polymorphism . If this newly invading inversion was subject to strong directional selection in its rise to its current frequency ( ∼46% ) —and this selection occurred in the recent enough past—a signature of selection on the level and frequency of nucleotide polymorphism should be evident . We used the sequence data collected from the 11 loci in and around the inversion to detect such a pattern . Levels of nucleotide diversity in the rearranged region ( calculated from the two separate samples of 2La and 2L+a chromosomes ) were lower than in collinear regions ( π: 1 . 11% versus 1 . 71% , Wilcoxon Rank-Sum test with 1-tail , p = 0 . 03; θ: 1 . 34% versus 2 . 14% , p < 0 . 02 ) , as expected if there has been recent directional selection in the inversion . However , contrary to the expectation of the selective sweep hypothesis , levels of diversity within the inversion are highest within ∼1 Mb of the proximal breakpoint and generally decline moving distally ( Table 1 ) . The one exception to this pattern of declining heterozygosity is high levels of polymorphism in the cpr34 gene in the distal cluster . A second prediction of the selective sweep hypothesis is that 2La chromosomes should contain less polymorphism than 2L+a chromosomes because of their more recent common ancestry . This pattern was not found: in fact , average levels of nucleotide diversity were slightly higher in 2La than in 2L+a arrangements , though the difference was only significant when diversity was estimated from the number of segregating sites ( π: 1 . 21% and 1 . 02% , Wilcoxon Signed-Rank Test , p < 0 . 20; θ: 1 . 59% and 1 . 09% , p < 0 . 01 ) . In addition , HKA tests [34] comparing A . gambiae 2La with A . quadriannulatus ( 2L+a ) , and A . gambiae 2L+a with A . arabiensis ( 2La ) across loci from rearranged and collinear regions were not significant for either comparison ( χ2 = 4 . 54 , p < 0 . 96 and χ2 = 9 . 82 , p < 0 . 40 , respectively ) . These test results indicate that there is not significant heterogeneity in levels of diversity relative to divergence between rearranged and collinear regions , consistent with the absence of recent hitchhiking on 2L and the lack of major differences in mutation rate between lineages . Two within-locus tests of deviation from the neutral-equilibrium model were conducted separately for 2La and 2L+a arrangements . Similar to previous sequence surveys of A . gambiae ( e . g . , [30] ) , Tajima's D statistic [35] was negative in most cases both inside and outside the rearranged region ( Table 2 ) , indicating an excess of low frequency SNPs ( single nucleotide polymorphisms ) consistent with a population expansion in A . gambiae [36] . None of the values of Tajima's D were significant , under equilibrium population histories or more realistic scenarios with expanding populations . However , four values of the R2 statistic [37] , also based on the site frequency spectrum , indicated a significant excess of low frequency polymorphisms relative to the neutral-equilibrium expectation . Also evident in measures of the site frequency spectrum—and consistent with the selective sweep hypothesis—were the more extreme values of both statistics in 2La-arrangement chromosomes . At seven of nine genes within the inversion , values of Tajima's D and R2 were lower ( indicating a greater excess of low frequency polymorphisms ) among 2La chromosomes . This result would be expected if the 2La arrangement rose in frequency quickly , though this explanation is somewhat undermined by the fact that Tajima's D and R2 are also lower at loci in collinear regions among individuals carrying the 2La arrangement . Although no clear footprint of a recent selective sweep or of balancing selection was found in the nucleotide sequence data , it may still be the case that selection is responsible for the maintenance of the proximal and distal clusters in association with inversion 2La . Multiple SNPs in both the proximal cluster and the distal cluster are in perfect linkage disequilibrium ( D′ = 1 ) with the inversion ( i . e . , they are fixed between inversion arrangements ) , even though they are quite distant from the breakpoints . The alternative to a selective explanation for their maintenance is that the observed linkage disequilibrium is an historical remnant of complete association dating from the time that the inversion entered the A . gambiae gene pool . Two lines of evidence suggest that this date is quite recent . First , A . gambiae itself is considered quite recently derived . Based on its strongly anthropophilic behavior and its dependence upon anthropogenic breeding sites , Coluzzi and colleagues [20] have argued that A . gambiae is the product of a speciation process originating in the central African rainforest and driven by human impact on the environment subsequent to the Neolithic revolution ∼10 , 000–12 , 000 years ago . Second , based on the assumption of a single introduction of 2La , we can derive an estimate for the age of 2La in A . gambiae that agrees fairly well with this time frame . After removing polymorphisms shared between arrangements and between species , we used nucleotide polymorphism data from the proximal breakpoint ( i . e . , endp ) to estimate that the E[TMRCA] of our sample of 2La chromosomes is ∼2 . 7 Ne generations ( where Ne is the effective population size ) . Microsatellite-based estimates of Ne of A . gambiae are reasonably consistent across Africa [38] . Values of Ne obtained from Cameroon based on the infinite alleles or stepwise mutation models of mutation , respectively , were 11 , 500 and 49 , 000 [38] . This corresponds to the introduction of 2La into A . gambiae ∼3 , 000–11 , 000 years ago , assuming 12 generations per year . Despite the relatively recent introduction of the 2La inversion into A . gambiae , we can distinguish between selective and neutral explanations for the maintenance of the inversion polymorphism by examining the amount of linkage disequilibrium expected between each of the clusters and their closest breakpoints given known rates of crossing-over . Using polymorphic microsatellite loci , Stump et al [30] estimated recombination rates on 2L from the backcross progeny of 2La/+a heterokaryotypes and as a control , from 2L+a homokaryotypes . They found that although recombination was at least 4× lower inside the inversion than in collinear regions , there were still appreciable levels of both gene conversion and crossing-over . From these data we estimate that the recombination fraction between the midpoint of the proximal cluster and the proximal breakpoint is r = 0 . 0012 , and that the fraction between the midpoint of the distal cluster and the distal breakpoint is r = 0 . 0168 . Given these estimates of recombination and Ne = 11 , 500–49 , 000 [38] , the quantity 4Ner is much greater than 1 for the regions in-between both clusters and their closest respective breakpoints . With 4Ner ≫ 1 , the only linkage disequilibrium expected in a population should be due to sampling variance [39]; we find that the observed value of the non-normalized linkage disequilibrium coefficient , D , is highly significantly different than 0 between polymorphic sites in either cluster and the inversion ( p = 1 . 08 × 10−11 for the smallest sample size of any locus in the proximal and distal clusters ) . This result strongly supports the conclusion that some form of natural selection must be maintaining the association between the individual clusters and the inversion , and therefore the inversion polymorphism itself . As an alternative way of considering the highly unlikely nature of the values of linkage disequilibrium observed , recall that disequilibrium declines as Dt = ( 1 − r ) tD0 , where Dt is the disequilibrium expected after t generations starting from an initial value of D0 . Given values of r ( see above ) and a starting value of D0 = 0 . 25 ( i . e . , complete linkage disequilibrium ) , we would expect Dt to be less than 0 . 001 between the proximal cluster and the inversion in 4 , 600 generations and less than 0 . 001 between the distal cluster and the inversion in 190 generations . These numbers of generations translate to an almost complete lack of linkage disequilibrium after 380 years for the proximal cluster and 16 years for the distal cluster . If our estimates for the date of introduction of the 2La polymorphism are within even an order of magnitude of the correct time , these results suggest that more than enough time has elapsed for the decay of disequilibrium between these highly diverged regions and the inversion itself .
There have been numerous models proposed to explain the maintenance of inversion polymorphisms ( reviewed in [2 , 40] ) . Perhaps the two most commonly cited are epistasis ( coadaptation ) among alleles within an inversion and overdominance of inversion heterokaryotypes [1] . Overdominance is an unlikely mechanism in this case . Multiple instances of stable geographic clines of 2La frequency along aridity gradients suggest that alternative arrangements are differentially adapted to dry and humid conditions , and that the cline results from a balance between migration and differential selection at opposite ends of an ecotone . This conclusion is reinforced by cyclical changes in 2La frequency associated with rainy and dry seasons each year . In addition , the overdominance hypothesis does not make clear predictions regarding linkage disequilibrium between loci within the inversion and the inversion itself , as a number of molecular mechanisms might be responsible for heterosis . In contrast , Dobzhansky's coadapted gene hypothesis makes clear predictions about the nature of epistasis among alleles within the inverted region and linkage disequilibrium between these alleles and the inversion . The observation in Drosophila of linkage disequilibrium between inversions and genes only loosely linked to breakpoints has lead previous researchers to suggest that epistasis for fitness was maintaining these inversion polymorphisms , though epistasis was not directly evaluated in these studies [12 , 13 , 41] . An alternative selective hypothesis for the maintenance of inversions that does not require epistasis is also consistent with these findings . As in the case presented here , all of the inversion polymorphisms in which epistasis has been invoked exist along stable clines [12 , 13 , 41] . Population structure at multiple loci under selection across the cline can generate linkage disequilibrium among loci , in a multi-locus analog of the Wahlund effect [42] . This occurs because local adaptation to different environments at multiple loci can lead to parallel clines in allele frequencies , and therefore nonrandom associations among alleles . Under this model , migration-selection balance at two or more loci maintains the inversion; epistasis among the locally adapted alleles is not required and therefore the requirements of this model are less stringent than for coadaptation [8] . For A . gambiae , clines from arid to humid environments across Africa offer an ideal opportunity for local adaptation in multiple traits ( see next section ) . Interbreeding among migrants carrying different genetic backgrounds on a collinear chromosome ( e . g . , humid- or arid-adapted ) would create recombinants bearing fewer humid-adapted ( arid-adapted ) alleles , resulting in lower overall fitness under humid ( arid ) conditions . However , inversions that capture all humid-adapted alleles preserve their association in the face of immigrant arid-adapted genes ( and vice versa ) . Thus , the inversion is maintained because it prevents recombination in the face of high levels of gene flow , as are observed in A . gambiae [43] . Further experiments will be needed to fully distinguish between the epistasis and local adaptation models for the maintenance of 2La . Our data predict that the proximal and distal clusters should contain at least some candidate genes that confer resistance to aridity on 2La ( and tolerance to humidity on 2L+a ) , though it is important to emphasize that additional candidates can occur outside of these clusters and possibly outside of the inversion itself . Within their estimated boundaries , a total of 210 genes have been predicted in both clusters . The challenge of identifying candidate genes within clusters is complicated by the fact that in many cases there is little evidence supporting gene predictions , with poor or nonexistent functional annotation ( Tables 1 and 2 ) . The effort is further complicated by an almost complete lack of information regarding the physiological and/or behavioral traits responsible for aridity tolerance conferred by 2La , which can include both desiccation resistance and resistance to heat stress . In the only published study of desiccation resistance and water balance in A . gambiae and A . arabiensis , a laboratory colony of A . arabiensis was significantly more resistant to desiccation than a colony of A . gambiae , due to higher initial body water content [44] . Metabolic rate , respiratory pattern , rate of water loss during desiccation , and water content at death were similar . As karyotype was not investigated nor controlled for during this study , these data are difficult to interpret with respect to the contribution of 2La; both colonies are known to be polymorphic for several inversions that have been associated with aridity in the field and the A . gambiae colony used was polymorphic for 2La . The same problem applies to the sole study of heat resistance that found A . arabiensis to be more heat tolerant than A . gambiae in a behavioral assay and stress test [45] . In the absence of more detailed guidance from empirical work , the most striking observation about gene content concerns the distal cluster , which contains the largest concentration of cuticle protein genes ( 40 ) in the A . gambiae genome , as well as three hsp83 genes encoding heat shock proteins . However , the cuticle proteins are not present in the epicuticle , the layer primarily responsible for water retention [46] . Thus their role—if any—in heat or desiccation resistance remains obscure . Substantial additional effort will be required to pinpoint the important genes and to understand their contributions to adaptive phenotypes . As alluded to above , 2La is not the final story on resistance to desiccation; other inversions on 2R are also implicated in this trait [3 , 27] . Future progress will depend upon controlling for karyotype differences . In the group of sister species known as the A . gambiae complex , there is a clear correlation between inversion polymorphism and involvement in malaria transmission [3] . The least polymorphic species are relatively restricted in their geographic distributions and are only locally important vectors or—in two cases—non-vectors . On the other hand , A . arabiensis and A . gambiae are counted among the most important vectors of human malaria worldwide . They carry abundant inversion polymorphism and are distributed across most of tropical Africa and its diverse landscapes . The impact of inversion 2La on the distribution of A . gambiae has been particularly profound . Once acquired from A . arabiensis , it helped A . gambiae to spread outside of the humid rainforest into arid savannas . Polymorphism for this and other inversions has enabled an already proficient malaria vector to occupy a vastly expanded species range , consequently expanding malaria transmission . Our results have laid the groundwork for the functional genomics study of 2La which will illuminate not only the genetic basis of adaptations inside inversions , but also aspects of vector behavior relevant to control .
All mosquitoes used in this study were field-collected . Collections of A . gambiae and A . arabiensis were performed between May and September of 2005 in the village of Tibati , Cameroon ( 6°28′N , 12°37′E ) by pyrethrum spray catch . A . gambiae s . l . were identified morphologically and the ovaries of half-gravid specimens dissected and fixed in Carnoy's solution ( 3:1 ethanol:glacial acetic acid ) . Sibling species and molecular forms M and S were identified using an rDNA assay [47] . Karyotyping was performed following standard protocol [48] . Inversion status of 2La was confirmed by a PCR diagnostic [49] . A . quadriannulatus specimens were collected in 1986 from southern Zimbabwe and kindly provided by F . Collins [50] . DNA was isolated from individual mosquitoes using the DNeasy Extraction Kit ( Qiagen ) . The concentration of eluted DNA for each specimen was determined by spectrophotometry using the Nanodrop-1000 ( Nanodrop Technologies ) . Fragmentation and labeling of 300 ng DNA from single specimens was achieved using random prime labeling in the presence of biotin-14-dCTP ( BioPrime DNA Labeling System , Invitrogen ) as described by J . Borevitz ( http://naturalsystems . uchicago . edu/naturalvariation/methods/BorevitzSFPMethods . pdf ) . After purification by ethanol precipitation , labeled products were resuspended in 100 μl ddH20 . Quality and yield ( estimated at ∼10 μg ) were checked by electrophoresis of a 5 μl aliquot through a 2 . 5% agarose gel . Most products were ∼50 bp long . The remaining 95 μl of labeled genomic DNA was hybridized to the Affymetrix Anopheles/Plasmodium GeneChip using standard protocols for eukaryotic cRNA hybridization . Hybridization and scanning of arrays was performed by the Center for Medical Genomics , Indiana University Medical School . All arrays were processed under identical experimental conditions on the same day . Cel files containing the raw probe intensity values were imported into Bioconductor ( http://www . bioconductor . org ) , an open source software project based on the R programming language ( http://www . r-project . org ) . Using the “affy” package , data quality was assessed to identify aberrant chips or spatial artifacts [51] . Approaches included examination of chip images of raw probe intensities at natural and log-scales , boxplot and histogram summaries of unprocessed log scale probe intensities for each array , and MA-plots . To visualize more subtle spatial artifacts , the affyPLM package was used to examine chip pseudo-images based on the probe level model ( PLM ) fit . Background adjustment and quantile normalization was performed using the Robust MultiArray Average ( RMA ) method without summarization by probeset [52] . Probe level data were exported as a comma separated value file for importation into Excel and are available from BJW upon request . Probes from the Anopheles/Plasmodium GeneChip have been mapped against the A . gambiae reference genome ( AgamP3 ) . To identify any probes with exact matches to multiple genomic locations or secondary one-off mismatches , a list of all probes and their genomic locations was obtained through VectorBase ( www . vectorbase . org ) [53] from K . Megy . A Perl script ( available from BJW upon request ) was used to parse probes with exact matches to unique locations; those with multiple exact matches or additional single base pair mismatches were excluded from further analysis . For each of the 151 , 213 probes retained a two-tailed t-test was performed to compare background-adjusted and normalized hybridization intensity values obtained from the five 2La arrays versus the five 2L+a arrays . Probes with p-values less than 0 . 01 were considered to contain SFPs between arrangements [14–16] . Overlapping significant probes were collapsed into one observation to control for nonindependence [16] . To test for overrepresentation of SFPs on 2L , we compared observed and expected numbers on 2L versus all other chromosomes combined , by a χ2 test . The expected number of SFPs in each category was calculated based on the genome-wide proportion of 0 . 89% as measured in this experiment . Similarly , overrepresentation of SFPs in the rearranged versus collinear part of 2L was tested by comparing observed and expected numbers given the 2L-specific proportion of 1 . 29% . An independent test of nonrandom SNP distribution on 2L that did not depend on prior information about the location of breakpoint sequences was implemented through a two-state HMM to identify differentiated versus homogenized regions along the arm . Transmission and emission probabilities of the HMM were estimated by expectation-maximization; hidden states were then inferred using the Viterbi algorithm in MATLAB ( The MathWorks , http://www . mathworks . com/ ) . To test for clustering of significant probes within the rearranged region , a sliding window analysis was performed with windows of 300 probes and a step-size of 20 probes . Each window was tested ( χ2 ) for an excess of significant probes compared to the number expected by chance . A Bonferroni correction for multiple tests was conducted using the effective number of independent tests according to the relationship n* = n ( 1 − ρ ) 2 , where n is the nominal number of tests conducted and ρ is the autocorrelation between successive test statistics [54 , 55] . A . gambiae GeneBuild AgamP3 . 4 incorporates manual annotations of genes predicted on 2L . Based on the manual models , primers targeting exons were designed using Primer3 [56] and custom synthesized ( Invitrogen ) . Primer sequences for each of the 11 exons studied and the corresponding VectorBase gene identifier is given in Table S1 . PCRs were carried out in a 50 μl reaction containing 200 μmol/l each dNTP , 2 . 5 mmol/l MgCl2 , 2 mmol/l Tris-HCl ( pH 8 . 4 ) , 5 mmol/l KCl , 10 pmol of each primer , 5 U Taq polymerase , and ∼10 ng of template DNA . Thermocycler ( MJ Research ) conditions were 94 °C for 2 min; 35 cycles of 94 °C for 30 s , 58 °C for 30 s , 72 °C for 1 min; a final elongation at 72 °C for 10 min; and a 0 °C hold . All 50μl of the resulting products were separated on a 1 . 25% agarose gel stained with ethidium bromide . Products were excised and purified using the Geneclean Spin Kit ( MP Biomedicals ) or QIAquick Gel Extraction Kit ( Qiagen ) . PCR products were directly sequenced on both strands using an Applied Biosystems 3730xl DNA Analyzer and BigDye Terminator version 3 . 1 chemistry as recommended by the manufacturer . Electropherograms were trimmed and visually inspected for SNPs and heterozygous indels using Seqman II ( DNASTAR , Madison , WI ) . Haplotypes at each locus were reconstructed from the genotypic sequencing data using the PHASE ( version 2 . 1 ) program , which implements a Bayesian statistical model for inferring haplotypes from population genotype data [57 , 58] . All default settings in PHASE were used except for tri-allelic and quad-allelic SNPs , for which the default assumption of stepwise mutation intended for microsatellite loci was relaxed . After two haplotypes were assigned to each specimen alignment was performed using ClustalX [59] . DnaSP version 4 . 10 . 9 was used to calculate standard polymorphism and divergence statistics and tests of neutrality [60] . Coalescent simulations of population expansion were conducted in ms [61] , with populations exponentially growing starting at 2 . 7 Ne generations in the past . Significance of FST values was based on 10 , 000 permutations conducted in Arlequin 3 . 11 [62]; significance of other values was determined from 10 , 000 coalescent simulations without recombination implemented in DnaSP [60] . A multilocus version of the HKA test of natural selection was implemented using HKA software developed and distributed by J . Hey ( http://lifesci . rutgers . edu/~heylab/HeylabSoftware . htm#HKA ) . Using maximum composite likelihood distances [63] , Neighbor-Joining gene trees were reconstructed in Mega4 [64] . To estimate the time to the most recent common ancestor of the 2La arrangement in A . gambiae , we used the expectation E[TMRCA] = 4Nef ( 1-ni−1 ) , which is based on the number of segregating sites unique to each of the inverted and standard classes [65] . This estimate assumes that the 2La arrangement entered the population and instantaneously reached its current frequency . Violation of this assumption makes E[TMRCA] a minimum estimate of the age of the inverted class .
All sequences mentioned in this paper have been deposited in the National Center for Biotechnology Information ( NCBI ) GenBank ( http://www . ncbi . nlm . nih . gov/sites/gquery ) under accession numbers EU097365 to EU097703 . | A chromosomal inversion occurs when part of the chromosome breaks , rotates 180 degrees , and rejoins the broken chromosome . The result is a chromosome carrying a segment whose gene order is reversed . Whereas the physical rearrangement itself may have no direct consequences on gene function , recombination between alleles in the rearranged and wild type segments is suppressed . If multiple alleles inside the inverted or original orientations are well adapted to contrasting environmental conditions , suppressed recombination provides a mechanism to keep beneficial allelic combinations from being shuffled between different genetic backgrounds . Working with wild populations of flies , Dobzhansky provided the first evidence that selection was key to maintaining inversion polymorphism . Subsequently , examples of inversion polymorphisms under selection in other organisms have been found , notably in the mosquito that transmits most cases of human malaria , Anopheles gambiae . However , the genes or gene regions conferring fitness advantages have yet to be discovered . In this study , the authors used modern genomics tools to map such regions in an inversion at an unprecedented level of detail , and show that these regions are likely to be responsible for the maintenance of the inversion polymorphism in natural populations . This study lays the groundwork for future efforts to identify the genes themselves and their role in adaptation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
]
| [
"evolutionary",
"biology",
"anopheles",
"genetics",
"and",
"genomics"
]
| 2007 | Localization of Candidate Regions Maintaining a Common Polymorphic Inversion (2La) in Anopheles gambiae |
Memories are assumed to be formed by sets of synapses changing their structural or functional performance . The efficacy of forming new memories declines with advancing age , but the synaptic changes underlying age-induced memory impairment remain poorly understood . Recently , we found spermidine feeding to specifically suppress age-dependent impairments in forming olfactory memories , providing a mean to search for synaptic changes involved in age-dependent memory impairment . Here , we show that a specific synaptic compartment , the presynaptic active zone ( AZ ) , increases the size of its ultrastructural elaboration and releases significantly more synaptic vesicles with advancing age . These age-induced AZ changes , however , were fully suppressed by spermidine feeding . A genetically enforced enlargement of AZ scaffolds ( four gene-copies of BRP ) impaired memory formation in young animals . Thus , in the Drosophila nervous system , aging AZs seem to steer towards the upper limit of their operational range , limiting synaptic plasticity and contributing to impairment of memory formation . Spermidine feeding suppresses age-dependent memory impairment by counteracting these age-dependent changes directly at the synapse .
Age-dependent memory impairment ( AMI ) , which is associated with both psychiatric and neurodegenerative disorders , starts in midlife and worsens with advancing age , suggesting that the greatest driving factor is age itself . The lack of effective treatments that prevent , halt , or reverse the condition is contributing to a diminishing quality of life for many senior citizens . Therefore , animal models that allow one to monitor physiological changes across their lifespan and to test for a causal character of age-induced changes might be helpful in exploring the mechanistic basis of AMI . D . melanogaster , with its short lifespan of around 60 d and advanced molecular genetic tools , provides an efficient experimental model to unravel mechanisms underlying AMI . Additionally , the olfactory nervous systems of insects and mammals exhibit many similarities , suggesting that the mechanisms for olfactory learning may be shared [1] . Moreover , aversive short- , intermediate- , and long-term olfactory memories have been found to be subject to age-induced decline in Drosophila , with an onset at about 10 d of age and plateaus at about 30 d of age [2–6] . Notably , we recently found a simple dietary supplementation of spermidine , a polyamine that specifically protects from AMI in Drosophila . External stimuli are believed to be represented in the brain as spatiotemporal patterns of neural activity within a set of neuronal connections . Changes in synaptic communication ( “plasticity” ) within certain neuron populations are meant to ultimately encode behavioral adaptations such as learning and memory . Thus , dysfunctioning of synaptic plasticity might well be relevant to age-dependent deterioration of learning and memory [7 , 8] . One of the fundamental problems of studying AMI , however , is the inability to differentiate causative changes from adaptive or protective changes . Moreover , the brain undergoes changes at multiple levels with advancing age , including alterations in circuits , individual neurons , and single synapses , further complicating the situation [8] . Nonetheless , recent work has linked AMI to subtle synaptic alterations in the hippocampus and other cortical brain areas , rather than to the loss of neurons [7 , 9] . At the same time , the age-associated modulation of molecular entities underlying learning and memory that define and change synapse function remain poorly understood . Therefore , we set out to determine the role of age-induced changes in the organization and function of synapses in AMI , using dietary supplementation with spermidine as a tool to identify synaptic changes that can potentially contribute to AMI . To accomplish this , we analyzed age-induced changes in the ultrastructural , molecular , and functional organization of synapses within the olfactory system of flies by comparing aged flies fed with normal food to aged flies fed with spermidine-supplemented food . We found that aging is associated with an increase in the average size of active zone ( AZ ) scaffolds , structures recently shown to scale with synaptic vesicle ( SV ) release . Consistent with this , optophysiological analysis showed that more SVs are released in response to natural odor stimuli in aged flies . Interestingly , these age-associated changes were suppressed by spermidine feeding , indicating that these changes might be causally relevant to AMI . In fact , genetic manipulation provoking an increase of T-bar size in young animals was sufficient to induce a premature decline in memory performance . We suggest that a cumulative increase in the size and function of presynaptic AZ scaffolds might reduce the operational range of synaptic plasticity processes , and thus , hamper the formation of new memories with age . Additionally , levels of postsynaptic neurotransmitter receptors and postsynaptic Ca2+ signals remained largely unaffected with age , suggesting that homeostatic adaptations might be involved in increasing the threshold for memory formation with advancing age .
In order to identify synaptic mechanisms plausibly contributing to AMI , we used opto-physiological assays to characterize overall neuronal responses in synaptic terminals of live intact flies . For these experiments , we focused on projection neuron ( PN ) to kenyon cell ( KC ) synapses within the mushroom body calyx of the olfactory system for two reasons: first , aversive olfactory learning involves coincidence detection of a conditioned stimulus ( odor ) with an unconditioned stimulus ( electric shock ) , causing changes in the odor-specific synaptic activity of second order PNs and third order mushroom body KCs [1 , 14]; second , the superficial position of the calyx within the fly brain enabled us to perform efficient optical analysis [15] , since sensor signals could be retrieved from discrete synaptic bouton areas . We started by expressing cytosolic GCamp3 . 0 in the PNs ( using GH146-Gal4 ) and found the basal expression of GCamp3 . 0 to remain largely unchanged with age ( S2A Fig ) . Next , we monitored the PN boutons for intracellular Ca2+ responses to two odors typically used for olfactory conditioning , 3-Octanol ( 3-Oct ) and 4-methylcyclehexanol ( MCH ) , through two-photon microscopy . Similar to our previous observations [3] , we found no significant difference in the amplitude or time course of cytosolic GCamp3 . 0 signals of young ( 3d ) and aged ( 30d ) animals ( S3 Fig ) . Thus , in the context of odor information processing , odor-evoked action potential frequency or presynaptic Ca2+ influx remained rather unaffected by the age of the animal . Next , we asked whether the release of SVs was altered with advancing age and analyzed the odor-driven SVs release . To this end , we used SynaptopHluorin ( SynpH ) , a pH-sensitive green fluorescent protein ( GFP ) fused to the luminal side of the SV membrane protein Synaptobrevin ( Syb ) [16] . SynpH is nonfluorescent at the acidic pH inside SVs; however , when SVs are released , SynpH is exposed to the neutral extracellular space , and the presynaptic terminal becomes brightly fluorescent . Following endocytosis , SVs become reacidified , and the cycle can start again [17] . SynpH was expressed within PNs , and the release of SVs in response to two odors was monitored , again , at PN-to-KC synapses ( Fig 1A–1H ) . We found a profound increase in the amplitude of SynpH signals in aged ( 30d ) animals when compared to young ( 3d ) flies ( Fig 1A–1H ) . In contrast , spermidine administration to 30d flies prevented this age-dependent increase of odor-induced SynpH signals ( 30dSpd; Fig 1A–1H ) . Alterations in the endocytotic clearance of newly released SVs might , per se , explain the increase in SynpH signals observed; however , the decay constants of the poststimulus SynpH signal remained essentially unchanged with aging ( S4 Fig ) , indicating that the endocytic clearance cannot be responsible for the difference in odor-driven SynpH signals observed in aged animals . In addition , neither the basal expression of SynpH before odor stimulation nor the maximal SynpH signal determined by high-molar KCl treatment showed systematic differences between young and aged cohorts ( S2B and S5 Figs ) . These experiments , thus , indicate that the exocytosis of SVs underlies the increase in SynpH response with advancing age . In addition to measuring the SV release at the PN presynaptic terminals within calyx , we also measured odor-evoked changes within the axonal projections of KCs within the mushroom body horizontal lobes by expressing SynpH using mb247-Gal4 . Though relative signals were smaller ( when signal was normalized to the whole mushroom body horizontal lobe ) , likely reflecting the well-documented sparse odor coding of KCs [18] , we still observed a substantially higher amplitude of SynpH signals in aged ( 30d ) than in young ( 3d ) flies , and , again , spermidine administration ( 30dSpd ) protected from this age-dependent increase ( S6 Fig ) . Thus , two major neuron populations of the olfactory system—PNs and KCs , showed an increase in odor-evoked fluorescence changes in response to odor stimuli , indicating higher release of SVs in aged animals . Since Ca2+ influx into presynaptic terminals was apparently not responsible for the profound age-induced increase in SV release , presynaptic mechanisms downstream of Ca2+ signaling might be involved . In order to address the molecular and cellular basis of this age-associated increase in SV release , we started by analyzing proteins directly associated with SVs: Synapsin , Syb , and Synaptotagmin-1 . Synapsin is a SV-associated phosphoprotein important for controlling the number of SVs available for release [19] , and Syb is a core component of SNARE complex that drives the exocytosis of SVs [20 , 21] . We found the levels of Synapsin as well as Syb to remain unchanged with advancing age ( comparing aged flies: 30-days old or 30d with young flies: 3d ) , regardless of spermidine feeding ( 30dSpd; Fig 2A–2D and S7 Fig ) . Synaptotagmin-1 is a vesicular protein with a central role as a Ca2+ sensor for SNARE-dependent SV fusion [22] . Synaptotagmin-1 decreased slightly with age , and feeding with spermidine had no discernable influence on this age-dependent change ( Fig 2E–2H ) , indicating that these moderate changes are seemingly not associated with AMI . The release of neurotransmitters is a sophisticated process that requires SVs to be in close vicinity to voltage-gated Ca2+ channels , and this precise spacing is orchestrated by interplay among several proteins that form the AZ scaffold [23 , 24] . In flies , the ELKS-family protein Bruchpilot ( BRP ) is an essential building block of the AZ scaffold and is needed to effectively cluster Ca2+ channels as well as regulate the release of SVs [25–28] . When whole-mount brains were stained for BRP using two different antibodies ( BRPNc82 and BRPN-term ) , we observed a substantial increase in the levels of BRP with advancing age ( Fig 2I , 2J , 2L and 2M ) . Similarly , Rim-binding protein ( RBP ) [29] , another structurally and functionally important component of the AZ scaffold , was found to be significantly increased in brains of 30d flies compared to 3d animals ( Fig 2I , 2J and 2N ) . Furthermore , flies analyzed at shorter intervals throughout their lifetime exhibited a progressive increase in the levels of both BRP and RBP ( S8 Fig ) . Notably , the age-dependent increase in BRP and RBP signals was suppressed in aged flies fed with spermidine ( 30dSpd; Fig 2I–2N ) . The staining efficacy could potentially be influenced by the sheer age of the tissue , e . g . , due to differences in antibody penetration . To rule this out , flies expressing a GFP-tagged genomic BRP construct ( rescuing the lethal brp null mutant [28] ) were aged on normal food or food supplemented with spermidine . We found the endogenous GFP signals to be significantly increased in 30d flies in comparison to 3d flies , while feeding with spermidine again prevented this age-related increase ( S9 Fig ) . Since the AZ scaffold has previously been reported to effectively cluster Ca2+ channels [26–28] , we asked whether the age-associated increase in levels of core AZ-proteins might influence synaptic levels of Ca2+ channels . To address this , we expressed a GFP-labeled genomic construct of α1 subunit Cacophony ( Cac ) , which is the only representative of the mammalian Cav2 . 1/2 . 2 family present in Drosophila [28] , and stained the flies for GFP and BRP . We found the levels of Cac ( quantified using an antibody against GFP ) to remain unchanged with aging ( S10 Fig ) . Besides its role in Ca2+ channel clustering , the AZ scaffold has been suggested to create a stereotypic arrangement that defines SV release slots by clustering SV release machinery [28] . In fact , the levels of Unc13 , a protein essential for priming SVs by rendering them fusion-competent [24] , were also increased in brains of 30d flies compared to 3d flies ( S11 Fig ) . Again , spermidine administration suppressed this age-dependent increase ( 30dSpd; S11 Fig ) . Taken together , our data suggest that synaptic levels of core AZ scaffold proteins , BRP and RBP , as well as the levels of critical release factor Unc-13 increased with advancing age . Next , we asked whether the increase of both BRP and RBP labeling in aged brains reflects an increase in the number of AZs or just the increase in local amounts of these proteins at individual AZs . To resolve this , we performed ultrastructural analysis on PN-to-KC synapses within the mushroom body calyx . In contrast to presynaptic terminals of KCs , presynaptic PN terminals within the calyx exhibit a well-defined morphology [30 , 31] , by which synapse types can be reliably identified in EM micrographs . Moreover , the superficiality of the calyx enabled us to perform stimulation emission depletion microscopy ( STED ) analysis ( see below ) . In order to allow for an unbiased quantification , we applied automated data collection to acquire more than a thousand transmission electron microscopic images covering nearly a whole calyx cross-section , which were then “stitched” together into a single high-magnification image ( see Materials and Methods ) . As described previously [30] , PN boutons could be easily identified , and light-colored boutons containing clear-core SVs were used for analysis . We recognized that plasma membranes between cellular elements were less aligned , with an increase in extracellular spacing between cellular elements , in aged ( 30d ) flies when compared to young ( 3d ) flies ( S12 Fig ) . Spermidine feeding appeared to substantially alleviate this age-related change ( S12 Fig ) . Driven by the finding that SV release is increased with age , we decided to analyze the AZs within PN boutons . We found aged animals ( 30d ) to display reduced numbers of AZs per unit bouton-area in comparison to 3d flies , with no apparent influence of spermidine feeding on this age-dependent decline ( Fig 3A–3E ) . The density of SVs in proximity to the AZ scaffold appeared unchanged in aged flies ( 30d as well as 30dSpd ) , when compared to young flies ( 3d; Fig 3F ) . Additionally , the number of SVs docked at the AZ plasma membrane appeared essentially unaltered with advancing age ( Fig 3G ) . The AZ scaffold exhibits an electron-dense structure in electron microscopy ( EM ) , and due to its T-shaped structure in Drosophila , this scaffold is often referred to as a T-bar [24 , 26 , 27] . We found the average size of the T-bars to be significantly increased in 30d animals in comparison to 3d flies ( Fig 4A–4D ) . Feeding flies with spermidine suppressed this age-induced increase in T-bar size ( 30dSpd; Fig 4A–4D ) . We have previously introduced STED in the analysis of AZ suborganization [26–28] . At peripheral neuromuscular synapses of Drosophila larvae , STED allowed us to unmask the “nano-architecture” of AZs where BRP and RBP organize a scaffold that provides slots for SV release and concentrates Ca2+ channels in the AZ center [28 , 29] . When planar AZs were imaged using the BRP C-terminal epitopes at neuromuscular synapses , they display a ring-shaped structure whose diameter correlated with the EM-derived physical size of individual T-bar/AZ scaffold [32] . We applied STED to PN-to-KC synapses of the calyx and found ring-like BRP structures at planar-oriented AZs ( S13 Fig ) . Subsequently , the analysis of these STED images revealed an increase in the ring diameter of BRP spots with advancing age , while spermidine treatment was able to suppress this age-associated increase ( Fig 4E–4H ) . Finally , we performed coimmuno-EM labeling against BRP and RBP on calycal slices . The number of gold particles positive for BRP as well as RBP was found to increase in aged flies ( 30d ) in comparison to both young ( 3d ) flies and aged flies fed with spermidine ( 30dSpd; Fig 4I–4N ) . Taken together , the morphological EM , immuno-EM , and STED analysis consistently show that aged animals display larger AZ scaffolds , plausibly due to an increase in local amounts of the critical scaffold components: BRP and RBP . Recent in vivo analysis of larval Drosophila neuromuscular junctions has shown that the local amounts of BRP at a given AZ scale directly with the probability of evoked SV release [33–37] . Consistent with these studies , we found SV release to increase and AZ scaffolds to enlarge with age , while importantly both these age-related changes were suppressed by dietary supplementation with spermidine . Therefore , we next wanted to determine the influence of these synaptic changes on olfactory memory formation . Presynaptic plasticity processes have been reported to be critical for forming olfactory associative memory in Drosophila [11–13] . Based on our findings , we suggest that the scale-up in the size and function of AZ scaffolds is likely to change the “operational range” of synaptic plasticity processes and thus change the threshold for memory formation . Thus , we wanted to test whether genetically provoking an artificial enlargement of AZ scaffolds , independent of the aging process , might affect memory formation . Since BRP is a major essential building block of the AZ scaffold in Drosophila [26–28 , 32] , we decided to increase the gene copy number of BRP from two to four copies by combining two additional genomic copies of brp [28] with two endogenous copies . As a result , BRP signals increased substantially in 3d flies expressing four-copy BRP ( 4xBRP ) when compared to 3d flies expressing two-copy BRP ( 2xBRP; Fig 5A–5E ) . Additionally , RBP levels also increased concomitantly with BRP ( Fig 5A–5D and 5F ) , consistent with the suggested role of BRP to operate as a “master molecule” in shaping the size ( and functional performance ) of the AZ scaffold [28 , 29 , 36] . In order to confirm that the increase in BRP levels resulted in an increase of the average size of AZ scaffolds , we took advantage of STED imaging . Again , a considerable increase in the ring diameter of BRP spots was observed in 2xBRP flies with advancing age ( Fig 5G–5K ) . Meanwhile , we found young flies ( 3d ) expressing 4xBRP to have increased BRP ring diameters when compared to age-matched control flies ( 2xBRP ) , and the ring diameter of BRP spots in 4xBRP flies remained rather unchanged with age ( Fig 5G–5K ) . Having created a genetic state wherein levels of AZ core scaffold proteins increased prematurely in young animals , we decided to investigate the influence of this manipulation on memory formation . Before doing so , however , we wanted to ascertain whether the innate behavior was affected in 4xBRP flies . Thus , we measured naïve odor response and shock reactivity and found 4xBRP flies to show odor avoidance and shock reactivity scores that were indistinguishable from 2xBRP age-matched control flies ( 2xBRP; S1 Table ) . Subsequently , we started by measuring short-term memory ( STM ) , and found 4xBRP flies to exhibit lower memory scores “already” at a young age ( 3d ) , and their memory scores declined only negligibly with age ( Fig 5L ) . In contrast , control flies ( 2xBRP ) exhibited normal AMI ( Fig 5L ) . As mentioned earlier , intermediate-term memory ( ITM ) has also been reported to decline with age [2–4] . Consistently , we found that 30d 2xBRP flies show substantially reduced ITM scores ( measured 3-h post-training ) when compared to 3-d 2xBRP flies ( Fig 5M ) . By contrast , the 4xBRP flies showed lower ITM scores at a young age ( 3 d ) and , again , the ITM scores did not decrease further in 30-d 4xBRP flies ( Fig 5M ) . In fact , the learning performance of 3-d 4xBRP flies was comparable to that of 30-d 2xBRP flies . Based on distinct genetic mutants and specific pharmacological sensitivities [2 , 4 , 38 , 39] , the ITM can be dissected into anesthesia-sensitive memory ( ASM ) and anesthesia-resistant memory ( ARM ) components . The ASM , unlike the ARM , has been shown to be strongly impaired with aging [3 , 4] . The ASM can be calculated by subtracting ARM scores , measured after amnestic cooling , from ITM . Consistent with previous studies [2–4 , 40] , we found ARM in 2xBRP and 4xBRP flies to remain relatively unaffected with age ( Fig 5M ) . In contrast , ASM was nearly absent in 30-d 2xBRP flies when compared to 3-d 2xBRP flies . Reaffirming our idea , we found the young ( 3-d ) 4xBRP flies to show lower ASM scores in comparison to age-matched control ( 2xBRP ) flies , while their ASM scores declined negligibly with age ( Fig 5M ) . These experiments indicate that a genetically provoked “up-scaling” of the average AZ scaffold size is sufficient to induce an “early” decline in memory , similar to AMI , which physiologically occurs over a time course of 20–30 d . A reduction in BRP levels , per se , might be expected to slow down the onset of AMI . To address this possibility , we removed a single gene copy of brp , and found BRP heterozygotes ( brp69/+ or 1xBRP ) to exhibit a considerable reduction in the levels of both BRP and RBP ( S14A–S14F Fig ) , indicating that our antibody stainings can detect subtle changes and reaffirming that BRP levels can directly steer the local amounts of other AZ components in the Drosophila brain . We found that 3d flies expressing only one BRP copy ( brp69/+ ) displayed memory scores comparable to 3d control flies ( 2xBRP ) ; however , these brp69/+ flies still exhibited a normally-occurring AMI ( 30d; S14G Fig ) . AZ scaffold-dependent control of neuronal plasticity is undoubtedly a complex process [24 , 41] , and other mechanisms , operating in parallel to modulations in the amounts of scaffold proteins , might well contribute to the pace and extent of AMI . Lysine-acetylation of BRP was recently identified as a major node to control the SV release at larval AZs [42 , 43] . In particular , the loss of histone deacetylase-6 ( HDAC6 ) was found to cause hyperacetylation of BRP and provoke a reduction in evoked SV release at AZs [43] . Interestingly , using immunoprecipitation followed by mass spectroscopic analysis , we found at least 13 lysine sites within BRP to be target for ( de ) acetylation , ( S15 Fig ) . Next , we asked whether loss of HDAC6 might affect memory . While the odor avoidance and shock reactivity were mainly unaffected by knockdown of hdac6 ( S1 Table ) , memory scores of both young and aged flies with pan-neuronal knockdown of hdac6 were higher than those of age-matched driver controls ( Fig 5N ) . These findings are consistent with the idea that driving down the AZs towards the lower limit of their operational range might facilitate memory formation in aged animals . Though any implications of acetylation of BRP or potentially other AZ scaffold proteins with respect to aging process still require extensive analysis , this result shows that BRP-directed modifications , reported to reduce SV release , can in fact increase the efficacy of memory formation in aged animals . Finally , we asked how the postsynaptic compartment might respond to these age-associated presynaptic structural and functional changes . To address this question , we used GCaMP3 . 0 fused to the postsynaptic protein Homer [15] and found the basal expression of Homer-GCamp3 . 0 to be largely unaffected with age ( S2C Fig ) . Moreover , the sensor was found to be effectively targeted to the postsynaptic density of the PN::KC synapses , as manifested by its specific enrichment within the postsynaptic specializations formed by claw-like dendritic endings of multiple KCs surrounding a single PN bouton ( Fig 6A ) . However , postsynaptic Ca2+ signals did not increase with age . Rather , a slight tendency towards a decrease of postsynaptic Ca2+ signals was observed in normally aged animals when compared to young controls ( Fig 6A–6H ) . At the same time , aged flies treated with spermidine ( 30dSpd ) produced signals more similar to untreated 3d-Homer-GCaMP3 . 0 flies than to untreated aged animals ( Fig 6A–6H ) . In order to be certain that Homer-GCamp3 . 0 signals were not saturated , we used high-molar KCl treatment to determine the maximal postsynaptic Ca2+ response . Unlike the odor-evoked maximum change in Homer-GCamp3 . 0 fluorescence of about 55% , KCl stimulation resulted in a substantially higher ΔF/F0 value of more than 300% ( S16 Fig ) , suggesting that sensor sensitivity was not a limiting factor for the postsynaptic Ca2+ signals . Meanwhile , when the cumulative postsynaptic Ca2+ activity was critically analyzed during the odor stimulation , we found that the Ca2+ responses reduced significantly in aged ( 30d ) flies relative to young flies , while the Ca2+ signals were comparable between young flies and spermidine-fed aged animals ( 30dSpd; S17 Fig ) . PNs provide cholinergic input to the KCs within the calyx [44] . We used a fusion of mushroom body-specific enhancer mb247 to the Dα7 subunit of the acetylcholine receptor ( mb247::Dα7GFP ) to explicitly visualize postsynaptic acetylcholine receptors . We showed previously that expression of Dα7-GFP from KCs localized specifically to the KC postsynaptic densities , where it closely matched the AZs of the PNs [45] . While we observed an age-related increase in BRP in 30d mb247::Dα7GFP flies in comparison to 3d mb247::Dα7GFP flies , the levels of Dα7 subunit ( quantified using an antibody against GFP fused to the α7 subunit of acetylcholine receptors ) did not change with age , and spermidine feeding had no effect on the level of the α7 subunit of acetylcholine receptors ( Fig 7A–7E ) . Similarly , when we stained for endogenous Drep2 , a postsynaptic scaffold protein that is known to express strongly within the postsynaptic densities of PN::KC synapses [46] , we also found Drep2 to remain unchanged with age ( S18 Fig ) . At first glance , the increase in release of SVs might be expected to translate into increased postsynaptic responses; however , ample evidence from various studies in different model organisms , including Drosophila , support the existence of homeostatic controls , allowing neurons to remain within a certain range of excitation [47 , 48] in order to avoid epileptic states and Ca2+-induced degeneration . In an attempt to directly examine the existence of such homeostatic controls , we wanted to determine whether an increase in the amount of depolarization required to trigger an action potential might influence the architecture of the apposed AZ scaffold . To achieve this , we used dORK1ΔC , a constitutively open K+ selective pore that causes hyperpolarization of neurons and subsequent inactivation of neuronal function [45 , 49] . dORK1ΔC was specifically expressed in the KCs , and presynaptic terminals of PNs within the calyx were analyzed for BRP levels ( Fig 7F–7I ) . Indeed , we found a substantial increase in the levels of BRP in the calyces of both 3d as well as 10d mb247>dORK1ΔC flies , when compared to age-matched controls ( Fig 7J ) . Thus , a drop in postsynaptic excitability can drive a homeostatic increase in presynaptic AZ scaffolds , leading to a potential increase in SV release at olfactory synapses—a finding similar to the one we found at aging synapses . Though the exact mechanisms allowing for homeostatic compensation of the elevated presynaptic release remain to be further worked out , it is tempting to speculate that homeostatic mechanisms coupling postsynaptic excitability to presynaptic release function might drive aging synapses towards the upper limit of their operational range and be critically involved in AMI ( see model in Fig 7K )
The aging process , causing progressive deterioration of an organism , is subject to a complex interplay of regulatory mechanisms . One of the primary aims of aging research is to use the understanding of this process to delay or prevent age-related pathologies , including AMI . We previously showed that restoration of polyamine levels by dietary supplementation with spermidine suppressed AMI in fruit flies [3] , providing us with a protective paradigm to identify candidate processes that might be functionally associated with AMI . As an insight towards the synaptic basis of AMI , we describe an age-induced increase in the levels of core AZ proteins , BRP , and RBP and of the functionally critical release factor Unc13 , together with a shift towards an enlargement of AZ scaffolds within the olfactory system . In addition , based on SynpH experiments , we observed a substantial increase in the release of SVs at aged synapses ( PN-to-KC and KC-to-mushroom body output neuron [MBON] synapses ) in response to odors used for learning experiments . Importantly , spermidine feeding was able to “protect” from both the functional and structural changes at aged AZs , arguing in favor of specific synaptic changes to be causally relevant for AMI . Indeed , installing 4xBRP not only increased the size of BRP rings in young flies , similar to those found in aged animals , but also provoked memory impairment in young flies . Notably , a reduction of BRP levels has previously been reported to affect ARM but not ASM [50] . Here , we report that an increase in BRP levels ( by changing the gene copy number of BRP from two to four copies ) severely affected ASM . These findings suggest that the two complementary forms of memory ( ARM and ASM ) might rely on the recruitment of distinct presynaptic “functional modules . ” The loss of brp has been shown to severely reduce release function in response to single low frequency , but not in response to high-frequency stimulation [27] , indicating that SV release at low-frequency stimulation might be particularly relevant for forming ARM , a memory component that develops gradually after training . On the other hand , mobilization of the SVs during high frequency stimulation has been suggested to be critical for formation of ASM [50] , a memory component that predominates early memory and decays with age . Thus , the increase in the size of the AZ scaffolds might potentially contribute directly to AMI by interfering with mechanisms facilitating SV availability in the course of forming ASM . Though the exact mechanisms underlying age-induced synaptic changes remain to be fully worked out , a reduction in autophagy-mediated protein degradation might well be involved [51–53] . Autophagy is a cellular digestion pathway that involves the sequestration of cytoplasmic components within a double-membrane vesicle called autophagosome , which fuses with lysosomes ( autolysosomes ) to degrade autophagic cargo by acidic hydrolases [52] . Interestingly , spermidine was shown to induce autophagy in several model systems , including rodent tissues and cultured human cells [51 , 54 , 55] . Moreover , amelioration of a-synuclein neurotoxicity due to spermidine administration was accompanied by autophagy induction [56] . Of note , we also found that spermidine feeding prevented accumulation of poly-ubiquitinated proteins by plausibly halting normally occurring age-induced decline of autophagic clearance [3 , 57] . The gene atg7 encodes an E1-like enzyme required for activation of both Atg8 and Atg12 , a step critical for the completion of the autophagic pathway [53] . We found that atg7-mutant flies ( atg7-/- ) exhibit reduced memory scores at a young age ( 3d ) , which declined further with age ( 20-d of age or 20d ) [3] . Concurrently , spermidine-mediated protection from memory impairment was eliminated in atg7−/− flies ( for both 3d- and 20d-flies ) [3 , 57] . Therefore , we wondered whether the decrease in the autophagic pathway might , per se , provoke increase of AZ scaffold components . When staining for BRP in atg7-mutant brains ( atg7-/- ) , we found a brain-wide increase in levels of BRP ( for both BRPNc82 and BRPN-term antibodies ) , and spermidine feeding was unable to prevent this age-related increase ( S19 Fig ) . The finding that spermidine feeding in atg7−/− flies neither blocked the increase in BRP levels ( S19 Fig ) nor suppressed memory impairment [3] suggests that the integrity of the autophagic system is crucial for the spermidine-mediated protection from age-associated increase in AZ scaffold components . Spermidine effects were recently shown to involve widespread changes of both nuclear and cytosolic protein acetylation [58 , 59] . In primary neurons , autophagosomes have previously been observed to form at the distal end of the axon , indicating compartmentalization and spatial regulation of autophagosome biogenesis [60 , 61] . More recently , autophagosomes were demonstrated to form directly near synapses and were found to be required for presynaptic assembly at developing synaptic terminals of Caenorhabditis elegans [62] . Moreover , the crucial release factor Unc13 was found to accumulate under conditions of defective endosomal microautophagy ( a specialized form of autophagy ) at developing neuromuscular synapses of Drosophila , suggesting Unc13 to be a substrate of this form of autophagy [63] . Interestingly , we have shown recently that the synaptic levels of Unc13-A isoform scale tightly with the levels of the BRP/RBP scaffold [64] . Thus , it is conceivable that some of the AZ proteins , whose levels increase with age ( BRP/RBP/Unc13 ) , might be direct substrates of “pre-synaptic autophagy , ” and that spermidine feeding might augment effective autophagic degradation of these proteins at aging synapses . We also observed a moderate decrease in synapse numbers in aging Drosophila brains , a phenotype that was unaffected by spermidine feeding . Our data compare favorably with studies in mammals . For example , loss of synapses in aged rodents has been reported in the dentate gyrus as well as the CA1 area of the brain [8 , 65 , 66] . Additionally , the “unitary” intracellular-evoked amplitude elicited by minimal stimulation protocols has been found to be greater in old than in young rodents [67] , suggesting that the “surviving synapses” are stronger [68] . It is of note that the induction threshold for long-term potentiation , considered to be a synaptic correlate of learning , has been reported to increase in aged rodents [10] . Similarly , an age-related increase in the amplitude of endplate potentials evoked has been reported at mouse neuromuscular synapses [69 , 70] . By contrast , a study at neuromuscular junctions of C . elegans revealed that aged motor neurons undergo a progressive reduction in synaptic transmission [71] . In flies , however , an age-related increase in the amplitude of the excitatory postsynaptic potential at adult neuromuscular junctions has been reported recently; this increase was suggested to tune the response of the homeostatic signaling system and establish a new homeostatic set point [72] . Collectively , these findings suggest that the dynamic range of synaptic plasticity may change with advancing age and , thus , contribute to AMI . Why would an increase in the odor-evoked SV release and ultrastructural size of AZ scaffolds impair the efficacy of forming new memories ? Synapses appear to display a “finite ceiling and floor” that define a synaptic operating range [73] . In rodents , the formation of new memories seems to drive synaptic strength to the upper limit of a fixed operating range , thereby creating an imbalance [73] . As a result , if the synapses are not returned to the midpoint of the synaptic modification range , then additional strengthening required for new memory formation might be blocked , and the system is driven to employ homeostatic compensatory mechanisms to balance the change [74] . In our experiments , we found dendritic Ca2+ signals and postsynaptic receptor levels to remain largely unchanged with age , suggesting the existence of homeostatic mechanisms that might allow the up-scaling of presynaptic release to be compensated by lowering the postsynaptic response to a given amount of neurotransmitter released . On the other hand , this upscaling of presynaptic structure and function might also be a homeostatic response to a reduction in postsynaptic excitability or Ca2+ homeostasis , steering retrograde enlargement of AZ scaffold and higher release of SVs ( Fig 7K ) . In fact , the influx of postsynaptic Ca2+ through glutamate receptors at the peripheral glutamatergic synapses of Drosophila has been reported to control presynaptic assembly by retrograde signalling [47 , 48 , 75] . While the exact nature of homeostatic controls connecting pre- with postsynaptic neurons in the olfactory system remains to be resolved , changes in plasma membrane excitability , a change in postsynaptic neurotransmitter sensitivity , or an increase in inhibitory GABAergic drive are obvious candidate processes . Taken together , we propose these synaptic changes steer the presynaptic AZs to function towards the upper limit of their operational range , making these synapses unable to react adequately to conditioning stimuli and provoke potentiation or depression of synapses in order to encode memory formation [11 , 12 , 76] . Sleep is widely believed to be critical for formation and consolidation of memories [77] . In sleep-deprived animals , neuronal circuits would exceed available space and/or saturate , thereby affecting an individual’s ability to learn [77] . Importantly , sleep deprivation has also been associated with widespread increases of BRP levels in the Drosophila brain [78] . Notably , we also observed a brain-wide increase in BRP levels in aged brains . It is tempting to speculate that both sleep deprivation and aging change the operational range over several synaptic relays and thereby affect memory formation—a topic that deserves further investigation in future . Taken together , our data show that upscaling of presynaptic structure and function contribute to an AMI in Drosophila . Furthermore , and restoration of polyamine levels prevents these age-associated alterations as well as AMI . Thus , spermidine feeding provides a unique opportunity to further the molecular and functional dissection of the mechanisms underlying AMI with the ultimate goal of restoring memory function in older humans .
All fly strains were reared under standard laboratory conditions [79] at 25°C and ≈70% humidity , with constant 12:12 h light/dark cycle . Flies from an isogenized w1118 strain were used as the wild-type control for all experiments . Flies carrying P ( acman ) cacGFP , P ( acman ) brp83GFP and P ( acman ) brp83 [28] and mb247::Dα7GFP [45] were described previously . The generation of UAS-homer-GCaMP3 . 0 flies are described elsewhere [15] . Briefly , cDNA of dhomer was amplified from w 1118 flies and inserted with a C-terminal linked GCaMP336 into pUAST . Both UAS-GCamp3 . 0 ( on the 3rd chromosome ) [80] and UAS-SynpH [81] were kindly provided by Gero Miesenböck . Atg7d14 and Atg7d77 flies were kind gifts from Thomas Neufeld [53] . In addition , mb247-Gal4 [82] and gh146-Gal4 [83] were used . As previously described [3] , the fly food was prepared according to Bloomington media recipe ( www . flystocks . bio . indiana . edu/Fly_Work/media-recipes/media-recipes . htm ) with minor modification , which was called Spd−or normal food . Spermidine ( Sigma Aldrich ) was prepared as a 2 M stock solution in sterile distilled water , aliquoted in single-use portions and stored at −20°C . After food had cooled down to 40°C , Spermidine was added to normal food to a final concentration of 1 mM or 5 mM Spd , and called Spd1mM+ or Spd5mM+ , respectively . Parental flies mated on either Spd−or Spd5mM+ food for all experiments , and their progeny were allowed to develop on the respective food . Flies used in all experiments were F1 progeny . The flies were collected once a day for aging , as a results-specific age indicated is day ± 24 h . Behavioral experiments were performed in dim red light at 25°C and 80% relative humidity with 3-Oct ( 1:150 dilution in mineral oil presented in a 14 mm cup ) and MCH ( 1:100 dilution in mineral oil presented in a 14 mm cup ) serving as olfactory cues , and 120V AC current serving as a behavioral reinforcer . Standard single-cycle olfactory associative memory was performed as previously described [3 , 4 , 46 , 84 , 85] , with minor modifications . Briefly , about 60–80 flies received one training session , during which they were exposed sequentially to one odor ( conditioned stimulus , CS+; 3-Oct or MCH ) paired with electric shock ( unconditioned stimulus , US ) and then to a second odor ( CS−; MCH or 3-Oct ) without US for 60 s with 30 s rest interval between each odor presentation . During testing , the flies were exposed simultaneously to the CS+ and CS− in a T-maze for 30 s . The conditioned odor avoidance was tested immediately after training for STM ( memory tested immediately after odor conditioning ) . Subsequently , flies were trapped in either T-maze arm , anesthetized , and counted . From this distribution , a performance index was calculated as the number of flies avoiding the shocked odor minus the number avoiding the nonshocked odor divided by the total number of flies and , finally , timed by 100 . A 50:50 distribution ( no learning ) yielded a PI of zero , and a 0:100 distribution away from the CS+ yielded a PI of 100 . A final performance index was calculated by the average of both reciprocal indices for the two odors . For ITM , flies were trained as described above , but tested 3 h after training . As a component of ITM , ARM was separated from ASM by cold-amnestic treatment , during which the trained flies were anesthetized 90 s on ice at 30 min before testing . In the end , ASM was calculated by subtracting the performance index of ARM from that of ITM for each training session on the same day , respectively . Brains were dissected in HL3 solution and fixed for 20 min at room temperature with 4% paraformaldehyde and 0 . 2% Glutaraldehyde in a buffer containing 50 mM Sodium Cacodylate and 50 mM NaCl at pH 7 . 5 . Afterwards , brains were washed twice in the buffer and dehydrated through a series of increasing alcohol concentrations . Samples were embedded in London-Resin ( LR ) -Gold resin by incubating them in Ethanol/LR-Gold 1:1 solution overnight at 4°C , followed by Ethanol/LR-Gold 1:5 solution for 4 h at room temperature . Thereafter , the samples were washed first with LR-Gold/0 . 2% Benzil overnight , a second time for 4 h , and then again overnight . Finally , the brains were placed in BEEM capsules covered with LR-Gold/0 . 2% Benzil resin and placed under a UV lamp at 4°C for 5 d to allow for polymerization of the resin . Following embedding , sections 70–80 nm , each , were cut using a Leica Ultracut E ultramicrotome equipped with a 2 mm diamond knife . Sections were collected on 100 mesh nickel grids ( Plano GmbH , Germany ) coated with 0 . 1% Pioloform resin and transferred to a buffer solution ( 20 mM Tris-HCl , 0 . 9% NaCl , pH 8 . 0 ) . Prior to staining , sections were blocked for 10 min with 0 . 04% BSA in buffer . Sections were incubated with the primary antibody ( guinea pig-anti RBPSH3II+III and rabbit-anti BRPlast200 , 1:500 dilution ) in blocking solution overnight at 4°C . After washing four times in buffer , the sections were incubated in buffer containing the secondary antibody ( goat anti-guinea pig 10 nm colloidal gold , goat anti-rabbit 5 nm colloidal gold British Biocell , 1:100 ) for 2–3 h at room temperature . Finally , the sections were washed four times in buffer and three times in distilled water . Contrast was enhanced by placing the grids in 2% uranyl acetate for 30 min , followed by washing three times with water and , afterwards , incubation in lead citrate for 2 min . Afterwards , the grids were washed three times with water and dried . Images were acquired on a FEI Tecnai Spirit , 120 kV transmission electron microscope equipped with a FEI 2K Eagle CCD camera . The following primary antibodies were used: MαBRPNc82 ( ref . 9 , 10; 1:100 ) , GPαBRPN-term ( 1:800 ) [25 , 27] , RbαRBPC-term ( 1:800 ) [29] , MαSynapsin ( 1:20 ) [89] , RatαSyb ( 1:100 ) [90] , RbαSynaptotagmin-1C-term ( 1:500 ) [91] , RbαGFP ( Molecular Probes; 1:500 ) , RbαDrep2C-term ( 1:500 ) [46] , RbαUnc13C-term ( 1:500 ) [63] , RbαBRPlast200 ( 1:500 ) , and GPαRBPSH3II+III ( 1:500 ) . The following secondary antibodies were used: GαM Alexa 488 ( Molecular Probes; 1:400 ) , GαR Alexa 488 ( Molecular Probes; 1:500 ) , GαGP Alexa 555 ( Invitrogen; 1:800 ) , GαM Cy3 ( Dianova; 1:500 ) , and GαR Cy5 ( Invitrogen; 1:400 ) . For Immunoprecipitation , BRPlast200 and IgG were used at final amount of 50 ug per 500 ul . For western blots , secondary antibody was used at a dilution 1:1 , 000 . Female 3d or 30d flies were briefly anesthetized on ice and immobilized in a small chamber under thin sticky tape . A small window was cut through the sticky tape and the cuticle of the head capsule using a splint of a razor blade . Trachea were carefully removed and the brain was covered with Ringer’s solution ( 5 mM HEPES , 130 mM NaCl , 5 mM KCl , 2 mM MgCl2 , 2 mM CaCl2 , pH = 7 . 3 ) . Imaging was performed using an LSM 7 MP two-photon microscope ( Carl Zeiss ) equipped with a mode-locked Ti-sapphire Chameleon Vision II laser ( Coherent ) , a 500–550 nm bandpass filter , and a Plan-Apochromat 20×1 . 0 NA water-immersion objective ( Carl Zeiss ) . A custom-built device to supply odorous air with a constant flow rate of 1 ml/s directly to the fly’s antennae was attached to the microscope . Odor stimulation ( MCH or 3-Oct , diluted 1:100 or 1:150 , respectively , in mineral oil or pure mineral oil ) was controlled using a custom-written LABVIEW program ( National instruments ) . GCamp3 . 0 , homer-GCaMP , and SynpH were excited at 920 nm and fluorescence monitored at an image acquisition rate of 5 Hz . The odorants were presented with a 20 s break between stimulation , and each fly was exposed to five to six repetitive experiments . The images were aligned to reduce small shifts in the X–Y direction using a custom written ImageJ plugin . The mean intensity within the region of interest of five images before stimulus onset was used as baseline fluorescence ( F0 ) . The difference in intensity ( ΔF ) was calculated by subtracting F0 from the fluorescence intensity value within the ROI of each image ( Fi ) and , subsequently , divided by the baseline fluorescence . ΔF/F0 values of three or more repetitions were averaged for each fly . Odor-induced fluorescence changes of SynpH were considered in calycal PN boutons showing ΔF/F0 values more than twice the standard deviation of the baseline fluorescence . The boutons with the five highest odor-induced ΔF/F0 amplitudes were considered for further analysis . We found SynpH to exhibit rapid photo-bleaching , therefore , bleaching correction was performed on its ΔF/F0 values . For this , first , ΔF/F0 values from the onset of the stimulus until the decay of the signal were removed and then the best least square fit was obtained using the remaining ΔF/F0 values ( second order polynomial decay function ) . Subsequently , this decay function was subtracted from the entire original ΔF/F0 curve , and the new modified data are the bleaching corrected data . Fluorescence emission of cytosolic GCamp was determined within specific boutons in the calyx that respond to the odor stimulus , and only the boutons showing ΔF/F0 values of more than 100% in four to five stimulations were averaged for each fly and considered for final analysis . Fluorescence changes of mb247-Gal4; UAS-homer-GCamp flies were averaged over the five most responsive microglomerular structures , as anatomically defined by basal fluorescence . False color-coded images were obtained by subtracting the image just before stimulus onset from the image at the maximum of the intensity difference ( i . e . , at 2 s after odor onset ) and divided by the baseline fluorescence . The KCl experiments were performed using a fluorescence microscope ( Zeiss ) equipped with a xenon lamp ( Lambda DG-4 , Sutter Instrument ) , a 14-bit CCD camera ( Coolsnap HQ , Photometrics ) and a 20 × NA = 1 water-immersion objective . Images were acquired at 5 Hz using Metafluor ( Visitron Systems ) . After recording some initial frames , KCl was added to the Ringer’s solution covering the fly brain ( final concentration 0 . 05 M ) . Fluorescence changes were determined in a circular region covering the calyx ( d = 20 μm ) , and background fluorescence determined outside the calyx was subtracted . For the identification of ( de ) acetylated residues of BRP , we did “conventional” protein extractions from Drosophila heads combined with BRP immunoprecipitations . The protocol could be divided into four main sections . 1 ) Precoupling of antibodies to matrix ( 50 ug antibody per reaction ) : 3 LoBind cups ( 2 ml; Eppendorf ) containing Affiprep Protein A matrix were prepared: 1 X 30ul for specific antibody , 1 X 30ul for IgG control , 1 X 60 ul for head extract preclearing . The cups were washed 3 X with 500 ul H-buffer ( 25 mM HEPES pH 8 . 3 ( NaOH ) , 150 mM NaCl , 1 mM MgCl2 , 1 mM EGTA , 10% Glycerol ) by inverting several times , followed by centrifugation 1 , 000 gmax ( 3 , 000 rpm ) for 1 min . 500 ul H-buffer ( + BRPlast200 or IgG ) per coupling was prepared . 500 ul antibody solution ( = 50 ug IgG ) was added per 30 ul washed Protein A-beads . Beads were incubated with antibody solution for 2 h on the wheel at 4°C . The Affiprep beads-antibody were collected by centrifugation for 3 min at 1 , 000 gmax . Affiprep beads-Antibody were washed 3 X by inverting tubes and 3 X for 10 min on wheel with IP buffer . 2 ) Homogenizing fly heads from stored fly heads [–80°C] . Fly heads were transferred with a clean spatula into 1 ml glass homogenizer . For 300 ul frozen fly heads , 300 ul Homogenization buffer ( without detergent ) was added , and heads were sheared at 900 rpm using an electronic overhead stirrer . Samples were collected in LoBind cups ( 2 ml; Eppendorf ) . 2 X 300 ul was added to rinse pestle and homogenizer ( Total volume in cups ~1 , 100–1 , 200 ul ) . Sodium-deoxycholic Acid ( DOC ) was added to a final concentration of 0 . 4% ( 28 ul of 10% stock spiked into homogenate ( 1:25 v/v ) ) . Triton X-100 was added to a final concentration of 1% ( 35 ul of 20% stock spiked into homogenate ( 1:20 v/v ) ) . The samples ( Homogenate ) were incubated for 60 min at 4°C at level 8 ( slow ) on wheel . 20 ul of homogenate was stored for SDS-PAGE analysis for monitoring antigen during extraction/pull-down procedure . Homogenate ( H ) was centrifuged for 15 min at 17 , 000 gmax . Supernatant ( yellow in color ) was transferred to a fresh LoBind cup . Centrifugation of S1 was repeated 4X to get rid of fat and remaining head debris . After final centrifugation step , remaining supernatant was diluted 1:1 with H-buffer ( without detergent ) . Total volume of Input was ~1 , 400 ul and of following composition: 25 mM Hepes pH 8 . 05 ( NaOH ) , 150 mM NaCl , 0 . 5 mM MgCl2 , 0 . 5 mM EGTA , 5% Glycerol , 0 . 2% DOC , 0 , 55% Triton X-100 . 3 ) Preclearing of fly head extract on Protein A-IgG beads: Diluted fly head extract was applied to preclearing beads and incubated for 60 min at 4°C while rotating on wheel . Precleared extract was separated by centrifugation for 3 min at 1 , 000 gmax . Supernatant ( IP input ) was recovered . 4 ) Precipitation: Precleared extract ( IP input ) was applied to antibody-bead matrix ( 600 ul to specific Antibody-beads , 600 ul to control IgGs ) and antibody–antigen binding was performed overnight at 4°C . Immunoprecipitates were collected by centrifugation at 1 , 000 gmax for 4 min at 4°C . Affiprep Beads-Antibody-Antigen were washed 3 X with a quick rinse followed by 2 X 20 min with 1 mL IP Buffer ( H-buffer + 0 . 5% Triton-X 100 + 0 . 2% Na-DOC ) . Affiprep Beads-Antibody-Antigen were resuspended in 1 , 000 ul IP buffer and transferred to a clean LoBind cup ( 2 ml; Eppendorf ) . Affiprep Beads-Antibody-Antigen were centrifuged , and most of the supernatant was removed ( without removing beads ) . 4 . ) Elution: For elution , 100 ul of 2X Laemmeli Buffer was added to Affiprep Beads-Antibody-Antigen and heated for 10 min at 95°C , 600 rpm , followed by centrifugation for 5 min at 1 , 000 gmax . Supernatant ( IP eluate ) was transferred into a fresh LoBind Cup ( 2 ml; Eppendorf ) . Immunoprecipitation was verified with western blot . For identification of ( de ) acetylated lysine residues in BRP , IP eluate was heated in SDS-PAGE loading buffer , reduced with 1 mM DTT ( Sigma‐Aldrich ) for 5 min at 95°C and alkylated using 5 . 5 mM iodoacetamide ( Sigma‐Aldrich ) for 30 min at 20°C . The protein mixtures were separated on 4%–12% gradient SDS‐PAGE ( NuPAGE , Invitrogen ) . The gel lanes were cut into ten equal slices , the proteins were in-gel digested with trypsin ( Promega ) [92] , and the resulting peptide mixtures were processed on STAGE tips [93] and analyzed by LC-MS/MS . Mass spectrometric ( MS ) measurements were performed on an LTQ Orbitrap XL mass spectrometer ( Thermo Fisher Scientific ) coupled to an Agilent 1200 nanoflow–HPLC ( Agilent Technologies GmbH , Waldbronn , Germany ) [94] . HPLC–column tips ( fused silica ) with 75 μm inner diameter ( New Objective , Woburn , MA , USA ) were self-packed with Reprosil–Pur 120 ODS–3 ( Dr . Maisch , Ammerbuch , Germany ) to a length of 20 cm . Samples were applied directly onto the column without a precolumn . A gradient of A ( 0 . 5% acetic acid ( high purity , LGC Promochem , Wesel , Germany ) in water and B ( 0 . 5% acetic acid in 80% acetonitrile ( LC–MS grade , Wako , Germany ) in water ) with increasing organic proportion was used for peptide separation ( loading of sample with 2% B; separation ramp: from 10%–30% B within 80 min ) . The flow rate was 250 nl/min and for sample application 500 nl/min . The mass spectrometer was operated in the data-dependent mode and switched automatically between MS ( maximum of 1 x 106 ions ) and MS/MS . Each MS scan was followed by a maximum of five MS/MS scans in the linear ion trap using normalized collision energy of 35% and a target value of 5 , 000 . Parent ions with a charge state from z = 1 and unassigned charge states were excluded for fragmentation . The mass range for MS was m/z = 370–2 , 000 . The resolution was set to 60 , 000 . MS parameters were as follows: spray voltage 2 . 3 kV; no sheath and auxiliary gas flow; ion transfer tube temperature 125°C . Data were analyzed in R v3 . 1 . 2 using the additional CRAN package dunn . test v1 . 2 . 2 . Asterisks are used in the figures to denote significance: * p < 0 . 05 , ** p < 0 . 01 , *** p < 0 . 001 , ns = not significant . Nonparametric methods were used because of the small sample sizes and because of failure of tests for normality for parts of the data ( Shapiro-Wilk test ) . Unless indicated otherwise , the different groups in each figure were first compared using the Kruskal-Wallis test , followed by Dunn’s test for posthoc multiple comparisons . Nonparametric tests were used in order to avoid being biased by outliers , which are represented by solid circles . All p-values that are reported have been subject to Bonferroni correction for the number of comparisons . Additional relevant information is indicated in the figure legends . The data for the behavioral studies were collected with the investigator blind to the genotypes , treatment , and age of genotypes . There was no blinding in the other experiments . The data were collected and processed side by side in randomized order for all experiments . In order to analyze the difference in Homer-GCaMP3 . 0 responses ( Fig 6 and S17 Fig ) , two-sided Kolmogorov-Smirnov tests were conducted in R , and the GCaMP3 responses only during odor stimulation and were compared . | Neurons communicate by sending impulses , in the form of secretion of neurotransmitters , across small spaces called synapses . It is these synapses that undergo structural and functional changes during formation and retrieval of memories . Though alterations in synaptic performance are believed to accompany aging , the causal relationship between age-dependent memory impairment and synaptic changes remains largely unknown . Using the fly Drosophila melanogaster as a model , we found that feeding them spermidine—a polyamine compound—suppresses age-induced decline in olfactory memory , providing us with a tool to further decipher mechanisms associated with age-dependent memory impairment . In this study , we investigated the relationship between synaptic changes and age-dependent memory impairment by studying the olfactory circuitry . We observed an age-related increase in the levels of the synaptic proteins Bruchpilot and Rim-binding protein , which caused an enlargement of the presynaptic active zone—the complex of proteins that mediate neurotransmitter release—and enhanced synaptic transmission . Interestingly , feeding of spermidine was sufficient to abolish these age-associated presynaptic changes , further emphasizing the relationship between presynaptic performance and age-dependent memory impairment . Furthermore , flies engineered to express an excess of the core active zone protein Bruchpilot showed a premature impairment in memory formation in young flies . Based on our data , aging plausibly steers the synapses towards the upper limit of their operational range , limiting synaptic plasticity and contributing to impairment of memory formation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"invertebrates",
"plant",
"anatomy",
"cognitive",
"neurology",
"medicine",
"and",
"health",
"sciences",
"fluorescence",
"imaging",
"nervous",
"system",
"electrophysiology",
"neuroscience",
"learning",
"and",
"memory",
"animals",
"animal",
"models",
"plant",
"science",
"model",
"organisms",
"drosophila",
"melanogaster",
"cognitive",
"neuroscience",
"immunoprecipitation",
"cognition",
"memory",
"nerve",
"fibers",
"flower",
"anatomy",
"drosophila",
"research",
"and",
"analysis",
"methods",
"imaging",
"techniques",
"animal",
"cells",
"presynaptic",
"terminals",
"cognitive",
"impairment",
"insects",
"precipitation",
"techniques",
"arthropoda",
"calyx",
"cellular",
"neuroscience",
"anatomy",
"synapses",
"cell",
"biology",
"physiology",
"neurology",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"cognitive",
"science",
"neurophysiology",
"organisms"
]
| 2016 | Spermidine Suppresses Age-Associated Memory Impairment by Preventing Adverse Increase of Presynaptic Active Zone Size and Release |
The rapid evolution of influenza viruses presents difficulties in maintaining the optimal efficiency of vaccines . Amino acid substitutions result in antigenic drift , a process whereby antisera raised in response to one virus have reduced effectiveness against future viruses . Interestingly , while amino acid substitutions occur at a relatively constant rate , the antigenic properties of H3 move in a discontinuous , step-wise manner . It is not clear why this punctuated evolution occurs , whether this represents simply the fact that some substitutions affect these properties more than others , or if this is indicative of a changing relationship between the virus and the host . In addition , the role of changing glycosylation of the haemagglutinin in these shifts in antigenic properties is unknown . We analysed the antigenic drift of HA1 from human influenza H3 using a model of sequence change that allows for variation in selective pressure at different locations in the sequence , as well as at different parts of the phylogenetic tree . We detect significant changes in selective pressure that occur preferentially during major changes in antigenic properties . Despite the large increase in glycosylation during the past 40 years , changes in glycosylation did not correlate either with changes in antigenic properties or with significantly more rapid changes in selective pressure . The locations that undergo changes in selective pressure are largely in places undergoing adaptive evolution , in antigenic locations , and in locations or near locations undergoing substitutions that characterise the change in antigenicity of the virus . Our results suggest that the relationship of the virus to the host changes with time , with the shifts in antigenic properties representing changes in this relationship . This suggests that the virus and host immune system are evolving different methods to counter each other . While we are able to characterise the rapid increase in glycosylation of the haemagglutinin during time in human influenza H3 , an increase not present in influenza in birds , this increase seems unrelated to the observed changes in antigenic properties .
The rapid evolution of influenza viruses presents difficulties in recognising and predicting current and future epidemiological threats . One of the major sources of information about possible future threats from influenza is a study of its history as an evolving pathogen . Analysing how the virus evolves to evade the immune response can provide insight into how the immune system has dealt with the virus in the past and how the virus may change in the future to evade elimination . Modelling of influenza evolution has focused on haemagglutinin ( HA ) , the membrane-bound glycoprotein present on the surface of the virus which is responsible for receptor-binding and membrane-fusion . Sixteen different HA subtypes in influenza A have been identified ( H1 to H16 ) of which H1 and H3 are currently circulating in human populations . For membrane fusion to occur , the HA precursor ( HA0 ) must be cleaved into two polypeptides , HA1 and HA2 , linked by a disulphide bond . Five canonical antigenic sites have been identified on the HA1 polypeptide of H3 [1] , [2] . Because HA1 is the principal target of antibody-mediated immunity [3] , it has a higher replacement rate for amino acids than HA2 [4] . The sequence evolution of HA1 results in antigenic drift as antigenic properties change with time . Amino acid substitutions result in changes in the ability of antibodies to neutralise the virus , either through interfering with antibody binding or changing some associated property ( e . g . receptor binding ) , so that antisera raised in response to one virus have reduced effectiveness against a future virus [4] . The amount of this reduction can be used as a measure of the difference between the antigenic properties of the two viruses . Interestingly , antigenic properties of H3 move in a discontinuous , step-wise manner [5] . For periods of two to five years , HA1 sequence evolution has a limited effect on virus-antibody interactions , so that antigenic drift is confined to a semi-well-defined group of sequence variants with similar antigenic properties , what has been referred to as an antigenic cluster . Periodically , however , sequence change results in significant change in antigenic properties , corresponding to a jump to a new antigenic cluster . The correlation between genetic distance and antigenic distance from the root cannot be explained by a linear relationship [5] . There are two possible , not mutually-exclusive interpretations of these irregular , punctuated changes . Firstly , it would be expected that changes in different locations would have different impacts on the antigenic properties , what has been described as the ‘influential sites’ model of antigenic change [6] . Some changes in the sequence will cause insignificant changes in antigenic properties , while other changes , near the antigenic or binding sites , would be more important . A constant rate of change of the sequence would result in punctuated changes in antigenic properties if relatively few locations had a very large effect on these properties [3] . ( In the data analysed by Smith and co-workers , jumps between antigenic clusters can result from a single amino-acid substitution [5] . ) Secondly , it might be that each antigenic cluster represents a particular manner of interaction between the virus and the host , such as nature or location of antibody binding . Changes in the amino acids at some locations would cause jumps between antigenic clusters , representing changes in this relationship . As a result , the effect of subsequent amino acid changes at other locations might be significantly different . In particular , Koelle et al . recently performed a simulation of the effect of such context-dependent interactions on the evolutionary dynamics of influenza , showing that it could re-create many observed epidemiological patterns [6] . One way to distinguish between these two possibilities is to look at the changing patterns of selective pressure . If changes in the antigenic cluster correspond to modifications in virus-host interactions , we would expect there to be corresponding changes in the selective pressure at different locations in the viral proteins . As the relative and absolute rates of amino acid substitution at these locations will depend upon the nature of the local selective pressure , we might be able to observe a change in the pattern of amino acid substitutions . These changes can be in the overall rate of substitution as well as in the nature of the substitutions accepted . One possible cause of these punctuated antigenic changes is the changing glycosylation state of the haemagglutinin . There has been a large increase in the number of predicted HA1 glycosylation sites from viruses isolated in 1968 to those circulating at present [3] , [7] . It would seem possible that the addition of these glycosylation sites represent a way for the virus to avoid the immune response through gross changes in the protein exterior . If these changes in glycosylation are related to changes in antigenic properties , we might expect correlations between the changes in glycosylation , changes in antigenic properties , and changes in the selective pressure at various locations in the protein . In addition to constructing phylogenetic trees , evolutionary theory can also be applied to a wide range of problems through hypothesis testing . These approaches can generate new insights into the forces of evolution that are shaping the protein sequence , and hence into the structure , function , and physiological context of the protein itself [8] . Competing models of sequence change can be applied to the data , and standard tools from statistics and information theory can be used to evaluate the evidence for specific behaviour . In this paper we are interested in addressing specific questions regarding haemagglutinin evolution . Does the selective pressure change during evolution , either in degree or nature ? Are the changes in selective pressure correlated with changes in antigenic properties or changes in glycosylation ? We develop a series of increasingly-complex models for the evolution of HA1 of human H3N2 viruses , at each stage inquiring whether we have statistical grounds for rejecting the simpler model . We start with a standard model where the rate of amino acid change at all locations is modelled by a single substitution matrix , allowing for heterogeneity of overall substitution rate following a Gamma distribution . We then develop a so-called ‘mixture model’ where different locations in the protein follow one of a set of possible substitution matrices , differing in overall substitution rate as well as different propensities for the various amino acids , but where we assume the substitution rates at any location is constant over the evolutionary process . The next model allows changes in the substitution rates during the evolution , corresponding to changing selective pressure . Because our mixture model includes substitution matrices that differ in their preference for different amino acids , we can detect changes in the nature of the selective pressure that do not correspond to changes in the magnitude or sign . Finally , we consider a more complicated model that considers that the alterations in selective pressure might preferentially occur on branches of the evolutionary tree corresponding either to changes in antigenic cluster or to changes in glycosylation . We find different substitution matrices describing different regions of the protein , indicating a range of selective pressures . We also find that these selective pressures change with time . More specifically , changes in selective pressure do not seem to occur at a constant rate throughout the tree . Rather , changes in selective pressure are found to occur more often during major changes in antigenic properties . This suggests that the movement between the antigenic clusters , as observed in Smith et al . [5] corresponds to changes in the nature of the interaction between virus and host . The locations that undergo changes in selective pressure are largely in places under positive selection , at or around the cluster-difference substitutions ( identified by Smith et al . ) , and in locations in canonical antigenic sites . Surprisingly , we do not observe a significant correlation between rapid changes in antigenic properties and changes in the predicted HA glycosylation state . Nor do we observe changes in glycosylation state during rapid changes in antigenic properties . This indicates that changes in glycosylation do not play a dominant role in the major changes of antigenic properties .
We developed four different evolutionary models . Model 1 , representing a standard optimised single substitution-model with Gamma-distributed rates , when applied to the haemagglutinin sequence data , yielded a log-likelihood of −4674 . 1 , for AIC = 10 , 492 . 2 . ( The number of parameters includes 209 model parameters and 363 adjustable branch lengths . ) We then developed a mixture-model ( Model 2 ) where there were a number of different substitution matrices representing different forms of selective pressure , defined by an overall substitution rate and different relative propensities for the various amino acids . Model 2 assumed that the selective pressure acting on every location was constant with time throughout the evolutionary process . The number of substitution matrices was optimised by minimising the AIC . The best performance was obtained with a mixture-model with four substitution matrices ( 271 adjustable model parameters ) , which achieved a log-likelihood value of –4339 . 2 for a substantially lower AIC = 9 , 946 . 4 . This indicates that an evolutionary model including qualitatively-different forms of selective pressure at different locations fits the data significantly better than a single substitution matrix with a Gamma distribution of rates . Allowing changes of selective pressure during the evolutionary process ( Model 3 ) increased the log-likelihood to −4319 . 8 . Model 2 ( no changes of selective pressure ) is nested in Model 3 , meaning we can use the likelihood ratio test to demonstrate that the extra parameters can be justified ( P<10−4 ) . We then tried a more elaborate model ( Model 4 ) , where the branches that involved changes of antigenic cluster had a greater amount of change of selective pressure . This increase of one additional adjustable parameter resulted in an increase in the log-likelihood to −4311 . 8 , indicating that this extra complexity is justified ( P<10−4 with the likelihood-ratio test ) , and that the increased rate of selective-pressure change between antigenic clusters is statistically-significant . We can then reject the null hypothesis that changes in selective pressure occur independently of jumps in antigenic properties , in favour of a model where changes in selective pressure occur preferentially coincident with such jumps . The rate of substitution-matrix change for the inter-cluster branches was γ = 0 . 77; that is , the extra amount of substitution matrix ( selective pressure ) changes is equivalent to what would be observed if the branch lengths corresponding to antigenic cluster transitions were increased by this amount . We performed ancestral reconstruction , and predicted glycosylation states of the various ancestral nodes . We restricted our analysis of glycosylation sites to the sites predicted on the ancestral nodes with probability >0 . 95 . We did not consider changes in glycosylation of the terminal sequences , as these might represent deleterious mutations , cannot be associated with changes in selective pressure with our model ( as there is no sequence evolution observed following the terminal sequence ) , and are independent of changes in antigenic cluster . Predicted glycosylation states are tabulated in Table S1 . We observe a sharp increase in the amount of predicted glycosylation sites from 6 sites ( HK68 ) to 11 ( FU02 ) , as shown in Figure 1 . Interestingly , there seems to be no correlation between changes in glycosylation and major changes in antigenic properties , with no transition between antigenic clusters corresponding to a difference in glycosylation . Conversely , some antigenic clusters contain a multiplicity of different glycosylation sites; WU95 viruses , for instance , contain between 7 and 10 glycosylation sites per subunit . This suggests that the rapid change of glycosylation is disjoint from the major changes in antigenic properties . We identified branches corresponding to a change in glycosylation . We then developed a substitution model ( Model 5 ) where the branches involving changes in glycosylation state were affected by an additional amount of selective-pressure change . There was a minimal change in log likelihood , indicating that there was not a significant observed correspondence between changes in glycosylation and changes in selective pressure , given the available data ( P = 0 . 8 ) . We find no evidence that changes in glycosylation correspond to significantly increased probability of changes in selective pressure . The results described in more detail below refer to Model 4 , unless specified otherwise . The amino-acid preferences of the four substitution matrices representing the four categories of selective pressure are represented in Figure 2 . The distribution of types of locations described by the different substitution matrices is shown in Figure 3A , while Figure 3B shows how the number of substitution matrix changes during the evolutionary process corresponds to various types of locations . In model 4 , locations can change between the different substitution matrices during the evolutionary process . The average rate of change between the different substitution matrices is tabulated in Table S2 . Substitution matrices one and two , representing 34% and 46% of all locations , respectively , are the slowest changing , with relative amino-acid substitution rates ( vk ) of 0 . 28 and 0 . 46 respectively . ( The substitution rates are normalised so that the substitution rate averaged over all sites is 1 . 0 ) The preferred set of amino acids is largely complementary between substitution matrices one and two , with substitution matrix one having an above-average hydrophobicity compared with substitution matrix two , although substitution matrix one contains an abundance of asparagine and glutamic acid while substitution matrix two has a propensity towards aromatic residues . As would be expected , the sequence changes of buried locations are preferentially described by the predominantly-hydrophobic substitution matrix one , while exposed locations not categorised as receptor-binding or in the canonical antigenic sites are preferentially described by the more hydrophilic substitution matrix two . Substitution matrices three and four , representing 9% and 11% of the locations in the protein , change relatively rapidly , with relative substitution rates vk equal to 2 . 93 and 3 . 90 . Substitution matrix three is biased towards positively charged amino acids ( arginine and lysine ) , while substitution matrix four has a bias towards small polar amino acids . Locations associated with the antigenic response–locations in the canonical antigenic sites and locations whose identity distinguishes the various antigenic clusters [5]–are predominantly described by these matrices , and also more likely to undergo changes in selective pressure . Loop regions are preferentially describable by substitution matrix four , in contrast to exposed coils . Receptor-binding sites are also more likely to correspond to these rapidly-evolving substitution matrices , corresponding to the high overlap between the receptor-binding and canonical antigenic sites . It is also possible that changes in the virus to prevent antibody neutralisation might involve modulating the receptor-binding properties directly , rather than inhibiting antibody binding . Table 1 shows the sites in the protein that undergo significant changes in selective pressure during transitions between antigenic clusters . As is shown , most of the changes occur in locations in the canonical antigenic sites , but there does not seem to be a preponderance of selective pressure changes in any particular site . Some of the locations undergoing selective pressure changes correspond to locations of the cluster-difference amino acid substitutions identified by Smith and co-workers [5] ( the K156Q substitution during the WU95→SY97 transition , and H75Q during the SY97→FU02 transition ) , while other changes in selective pressure occur near the cluster-difference substitutions ( e . g . location 157 near the G158E substitution during the EN72→VI75 transition ) . There is evidence for a change in selective pressure at location 124 during the TX77→BK79 transition , with a G124D cluster-difference substitution occurring during the subsequent BK79→SI87 transition . Many changes in selective pressure , however , occur in locations that are not directly associated with cluster-difference substitutions . Haemagglutinin must fulfil a number of functional requirements . Changes in antigenic properties might be correlated with adjustments of other properties , such as receptor binding , which could then be associated with changes in selective pressure not directly associated with the antigenic response . Finally , there might be compensatory changes due to , for instance , the effect of some substitutions on thermodynamic stability . The locations of different substitution matrices and changes in substitution matrix , compared with the canonical antigenic sites and those determined as being under positive selection are illustrated in Figures S2 and S3 , respectively . We observe rapid substitutions , as well as rapid changes of substitution matrix , in the central exposed ‘pore’ at the top of the protein . These locations are not identified either as being under positive selection , or as changing during changes in antigenic property , although they are centrally-located and surrounded by such locations . Changes involved in cluster transitions BE92→WU95 and WU95→SY97 are shown in Figures S4 and S5 , respectively .
There is evidence of changes in selective pressure during HA evolution . For instance , Wolf et al . recently observed transient ‘adaptive bursts’ characterised by positive selection occurring in epitopic regions [9] . In between these bursts there is little evidence for positive selection , and newly-emergent lineages only slowly replace existent lineages . There has also been evidence for non-transient shifts of selective pressure . For instance , while changes in the 18 locations identified by Bush et al . as under positive selection from 1983 to 1997 [10] seemed to be correlated with the subsequent phylogenetic trajectories [11] and changes in antigenic properties [5] during this same time period , changes in these 18 locations over a longer time-range were only weakly correlated with changes in antigenic properties [5] , [12] . A study of sequence change and sequence variability suggests that antigenic drift involves changes in a local region , but that the location of this region varied from transition to transition [13] . All of this suggests that positive selection is a feature of influenza evolution , but that the locations undergoing positive selection may change and new antigenic sites may emerge . As described in the introduction , the punctuated nature of the antigenic changes can be explained if different locations had different impacts on antigenic properties , and cluster-changes corresponded to changes at the more critical locations . In this case , the selective pressure might still be relatively constant or change in a way not correlated with the changes in antigenic properties . Alternatively , jumps in antigenic properties might represent change in the mechanism of immune-avoidance for the virus or changes in the antibody response . This latter alternative has been recently simulated with an epidemiological model [6] . The majority of substitutions occur within a cluster as populations evolve within a set of sequences with similar antigenic properties . These changes progress until a single or set of ( rare ) mutations cause a jump to a new antigenic variant with higher fitness . This sequence and its descendents replace the old cluster , resulting in the collapse of the population to a new single lineage that undergoes a new cycle of diversification . Our model provides evidence for the second explanation , that the antigenic clusters correspond to changing relationships in the ‘arms-race’ between influenza and the immune system , resulting in significant changes in selective pressure at different locations in the protein . These changes in selective pressure are quite rapid , corresponding to the amount of selective pressure change that would occur in a branch of length 0 . 7 , while the branch lengths for the transitions are on the order of 0 . 01 to 0 . 03: this represents a 20- to 70-times increase in the rate of changes of selective pressure . Consistent with this model , the changes in selective pressure occur predominantly in the canonical antigenic sites . These changes also occur at locations occupied by different amino acids in the different antigenic clusters [5] . It is important to note that these are not necessarily cluster-defining changes , in that some of these changes might have occurred independently of any changes in antigenic properties . Still , there is a strong correlation between sites undergoing such amino acid changes between antigenic clusters and locations where there are corresponding changes in selective pressure . There is also a significant tendency for changes in selective pressure in the regions surrounding the cluster-difference changes . Interestingly , we cannot detect a significant increase in the rate of substitution matrix change during changes in glycosylation . As is clear from Figures S1 and 1 , we also observe no correlation between changes in glycosylation and changes in antigenic clusters . None of the cluster-changing transitions involve a change in glycosylation site; conversely , many single antigenic clusters contain different HA with a wide range of different glycosylation states . This result is surprising , given the experimental evidence that glycosylation can reduce antibody binding [2] , [14]–[17] , although it is important to note that significant changes in antigenic properties can occur within the antigenic clusters . A similar analysis of glycosylation changes in H9 evolution in birds , thought to represent a situation of viral ‘stasis’ in a natural host , do not demonstrate any significant increase in glycosylation state ( data not shown ) . Similarly , the glycosylation state of H1 in humans does not show a substantial increase , with the number of glycosylation sites fluctuating between about 8 and 10 per subunit ( data not shown ) . The amount of glycosylation may represent a balance between antibody shielding and other requirements such as the need to modulate receptor affinity [18]–[20] and avoid the innate immune response; increased haemagglutinin glycosylation results in reduced virulence in mice due to virus binding by collagenous lectins [21] . Reduced binding of influenza viruses by this mechanism in humans might alter the balance towards increased glycosylation . Another possible explanation is that glycosylation-induced antigenic changes that might occur in humans would not be detected in ferrets , and thus do not show up in the antigenic property analysis of Smith et al . It is known , for instance , that humans contain a significant number of antibodies for galactose compared with ferrets [22] . It is not clear how this would explain the absence of correlation between glycosylation changes and changes in selective pressure . We note that we are examining changes in the predicted , rather than observed , glycosylation state . It is likely that a large fraction of these locations are , in fact glycosylated . The crystal structure of haemagglutinin from H3N2 A/Aichi/2/1968 ( PDB designation 5HMG ) is predicted to have six glycosylation sites per subunit , four of which are observed in the structure [23]; the remaining two might have been lost through protein expression , purification , or crystallisation . Furthermore , we consider it unlikely that the inaccuracy of the predictions is responsible for the lack of correlation between antigenic changes and glycosylation changes , as there is no reason to believe that there are significant numbers of predicted glycosylation sites that change their occupancy during changes in antigenic cluster while there are no sites that change their predicted state . Similarly , it is difficult to imagine that there is a correlation between undetected changes in occupancy of these predicted sites that correspond to increased changes in selective pressure when no such correlation is observed with changes in the predicted sites . We find strong support for a model where the selective pressure changes preferentially during transitions between antigenic clusters . This suggests that evolution of human H3 consists of periods of amino acid variation according to a relatively constant set of rules , interspersed with periods where the rules governing variation change . These issues have important consequences for the predictability of antigenic drift . If the selective pressure at different locations in the protein are relatively constant , we could directly extrapolate future changes from past changes , an assumption explicit in previous analyses [11] . If , however , changes in antigenic properties are associated with changes in virus-immune system interactions , we might have to model changes in this relationship in order to perform reasonable extrapolations , as important sequence changes during one interval of antigenic drift might not be the same as ones that are important during other intervals . In addition to modelling how amino acids change during time , we also may need to develop models for how the selective pressure changes . These results also suggest that the notion of canonical ‘antigenic sites’ might be overly simplistic . It appears that there are a wide range of different locations with different propensities towards antibody recognition , and that the specific haemagglutinin locations so targeted may change with time . If so , the distinction between antigenic and non-antigenic sites may be subtle and time-dependent .
As described above , we develop a series of increasingly complex models . Each increase in complexity , if justified by the data , demonstrates a simplifying assumption that can be rejected , providing increased understanding of the nature of the evolutionary process in influenza . Early simple evolutionary models , that assume that the rate of substitutions at all locations in all proteins at all times followed the same substitution matrix , have been gradually supplemented by mixture models that allow differences in the absolute substitution rates [24] , relative substitution rates at different locations [25]–[28] , and differences in the substitution rates at different times [29] , [30] . Each component of the mixture model , represented by a distinct substitution matrix , reflects a different degree or form of selective pressure . In the simplest models ( such as Gamma-distributed rate classes ) , we can consider different components as having different magnitudes of selective pressure , resulting in different absolute substitution rates . In the mixture models considered here , we allow for differences in the magnitude of the selective pressure as well as differences in the preferences for the different types of locations for the various amino acids . For instance , one component may model the inside of the protein , and so have a bias towards hydrophobic amino acids . Details of the various models are described in Protocol S1 . ( For an overview of standard approaches to evolutionary modelling , see e . g . [31] . ) Model 1 involves a standard single substitution matrix with Gamma-distributed rate variation [24] . In Model 2 , we consider that different locations in the protein follow , or ‘are assigned’ , to one of a number of different possible substitution matrices [25]–[28] . We do not initially know which sites belong to which substitution matrix . Instead , each substitution matrix k has a specified a priori probability P ( k ) of representing any given site in the protein at any time . ( As all sites must belong to some substitution matrix , ) . The different substitution matrices are characterised by an overall substitution rate νk , the relative frequencies for the twenty diverse amino acids {πi , k} , and a symmetric rate parameter matrix Si , j ( Si , j = Sj , i ) that is optimised over the entire dataset and is the same for all substitution matrices . The overall substitution rates are normalised so . Model 3 includes a rate at which a substitution matrix describing any given location can change to another during the evolutionary time , representing variations in the selective pressure on the protein over time . The various parameters for the substitution matrix model without changes in selective pressure are {Sij , P ( k ) , νk , πi , k} . Allowing changes in substitution matrix adds {Zkl} , a new symmetric matrix ( Zk , l = Zl , k ) adding an additional Nk ( Nk−1 ) /2 parameters for Nk substitution matrices . Models 4 and 5 consider the possibility that the rate of change of selective pressure , that is , the rate at which a single location changes from one substitution matrix to another , might depend upon the specific branch of the tree , depending , for instance , according to whether that branch involved a change in antigenic properties ( Model 4 ) or glycosylation state ( Model 5 ) . In these cases , we consider a model where these specific branches are subject to an additional substitution-model change matrix which only includes substitution matrix change but no additional changes of amino acid . We can then use the likelihood ratio test to see if the resulting improvement in the log likelihood justifies the addition of this additional parameter . To evaluate the models we have used the dataset of Smith et . al . [5] which contains 254 Human H3 HA1 sequences sampled from 1968 to 2003 . An avian H3 sequence ( A/Duck/Hokkaido/33/80 , M16739 ) was used as an outgroup to root the tree . Sequences were extracted from the Influenza Sequence Database [32] . The Maximum Likelihood phylogenetic tree was derived using PHYML [33] with the WAG substitution model [34] and a Gamma-distributed rate [35] . Various parts of the tree were assigned to different antigenic clusters according to the designations of Smith et al [5] . The listings of these antigenic clusters as well as the abbreviations used in the text are in the legend for Figure 1 . After the computation of the phylogenetic tree , the parameters of the model were optimised to maximise the log-likelihood , using software available from the authors . The probability of the different substitution matrices and amino acids at each location in the protein for each ancestral state were calculated using standard maximum-likelihood ancestral reconstruction methods [36] , [37] , as were the probability of changes in selection pressure . Ancestral glycosylation states were determined by searching for locations containing the sequence Asn-Xaa-Ser/Thr with probability >0 . 95 . Homology models of a representative set of ML ancestral sequences were made with SwissModel [38] based on the 1MQN structure [39] . When glycosylation states were predicted by the GlyProt server [40] , all potential locations were predicted to be glycosylated . We are often confronted with the choice of one model or another , of varying degrees of complexity and resulting fit to the sequence data . The relative fit of two different models is quantified by the ratio of their likelihoods ( that is , the probability that the observed data would be generated by the model ) , or equivalently , the magnitude of the change in log-likelihood . In some cases , these models are ‘nested’ , that is , one model ( A ) is a restricted form of model ( B ) , in which case we can use the likelihood ratio test to see if the added complexity is justified by the resulting increase in log-likelihood [41] . We cannot use the likelihood ratio test to evaluate the performance of non-nested models . Instead , we use the Akaike Information Criterion ( AIC ) [42] , which is defined as AIC = 2Np−2Λ , where Np is the number of adjustable parameters and Λ is the log-likelihood . The preferred model is that which minimises the resulting AIC . According to this criterion , a more complex model is only justified when it causes an increase in the log likelihood greater than the number of additional parameters . | H3N2-type influenza is responsible for widespread disease and significant mortality . The virus evolves rapidly , changing its antigenic properties , allowing it to escape clearance by the immune response as well as complicating the maintenance of vaccine effectiveness . Part of this evolution has been the rapid increase in glycosylation , an increase not observed either in H9 evolution in birds or in H1 evolution in humans . It has been observed that the antigenic properties change in a punctuated , discontinuous manner . This could be either because some mutations are more significant than others , or it could mean that the antigenic changes correspond to adjustments in the antagonistic relationship between virus and host . By studying the sequence evolution of the H3 haemagglutinin , we can demonstrate that the selective pressure acting on the virus protein changes with time , and that these changes are especially rapid during changes in antigenic properties . This indicates that the antigenic changes correspond to modifications in the virus–host relationship . Surprisingly , neither the changes in selective pressure nor the changes in antigenic properties correspond to changes in glycosylation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
]
| [
"computational",
"biology/evolutionary",
"modeling",
"infectious",
"diseases/viral",
"infections",
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"virology/virus",
"evolution",
"and",
"symbiosis"
]
| 2008 | Changing Selective Pressure during Antigenic Changes in Human Influenza H3 |
Hepatitis B is a DNA virus that infects liver cells and can cause both acute and chronic disease . It is believed that both viral and host factors are responsible for determining whether the infection is cleared or becomes chronic . Here we investigate the mechanism of protection by developing a mathematical model of the antibody response following hepatitis B virus ( HBV ) infection . We fitted the model to data from seven infected adults identified during acute infection and determined the ability of the virus to escape neutralization through overproduction of non-infectious subviral particles , which have HBs proteins on their surface , but do not contain nucleocapsid protein and viral nucleic acids . We showed that viral clearance can be achieved for high anti-HBV antibody levels , as in vaccinated individuals , when: ( 1 ) the rate of synthesis of hepatitis B subviral particles is slow; ( 2 ) the rate of synthesis of hepatitis B subviral particles is high but either anti-HBV antibody production is fast , the antibody affinity is high , or the levels of pre-existent HBV-specific antibody at the time of infection are high , as could be attained by vaccination . We further showed that viral clearance can be achieved for low equilibrium anti-HBV antibody levels , as in unvaccinated individuals , when a strong cellular immune response controls early infection .
Infection with hepatitis B virus ( HBV ) results in acute hepatitis followed by recovery in 85%–95% of human adults [1] . Recovery occurs when the organism mounts adequate immune responses against the virus . Such responses include production of protective , neutralizing antibodies against HBV surface antigen ( HBsAg ) [2] , [3] , activation of strong and diversified CD4 and CD8 T-cells [4] , [2] , expression of antiviral cytokines in the liver , such as gamma interferon and tumor necrosis factor alpha [5] , [6] , [7] , [8] , and generation of cells that are protected from reinfection [9] , [10] . In contrast , progression to chronic HBV infection occurs predominantly in immuno-compromised adults and in unvaccinated infants [11] . Such individuals exhibit weak and inefficient humoral and cellular immune responses , resulting in continual virus replication and HBV surface antigenemia [12] , [4] . Little is known about the relative contributions of different arms of the immune system , especially the roles of neutralizing antibodies in the onset and outcome of infection . The antibody response to HBV infection is difficult to study experimentally . Free antibody to surface antigen is not detected until after the resolution of HBV infection [13] . However , circulating immune complexes containing antibody and HBsAg are found in both acute and chronic HBV infections , suggesting that antibodies are produced much sooner than detected , and that they might play a role in the pathology of the disease [14] , [15] , [16] , [17] . HBsAg-specific antibodies have neutralizing properties and mediate protective immunity [16] . Infection with hepatitis B virus results in the synthesis of a large number , probably of at least 1 , 000-fold , of “subviral particles” ( SVPs ) in relation to HBV particles [1] , [18] . SVPs , which are produced by HBV infected cells , are particles that have HBV proteins on their surface , but do not contain nucleocapsid protein and viral nucleic acids and hence are non-infectious [19] . They exist in two main forms: spheres 25 nm in diameter and filaments 22 nm in diameter with variable lengths [20] , [21] , [22] , [23] . The reasons for their overproduction and their contribution to HBV pathogenesis is still under investigation [24] . SVPs may influence the way the host reacts to HBV infection . They may induce tolerance during perinatal infection , thus delaying the rise of neutralizing antibodies . Additionally , the excess of subviral particles can serve as a decoy by adsorbing neutralizing antibodies and therefore delay the clearance of infection . In this paper , we aim to determine quantitative features of the antibody responses to virus and subviral particles following HBV infection . We build on basic chronic virus infection models [25] , [26] , [27] , [28] , [29] , [30] , [31] and determine the antibody characteristics that explain both the high peak and eventual viral clearance observed during acute hepatitis B infections [8] . We show that antibody responses can lead to viral clearance when the anti-HBV levels are high , as in vaccinated patients , and: ( 1 ) the rate of synthesis of hepatitis B subviral particles is slow; ( 2 ) the rate of synthesis of hepatitis B subviral particles is high but either anti-HBV antibody production is fast , the antibodies have high affinity , or the levels of pre-existent HBV-specific antibody at the time of infection are high . For lower anti-HBV antibody levels , as in unvaccinated patients , both cellular and humoral responses are needed in concert to clear acute HBV infection with the CD8 T cells controlling the initial burst of replication and the antibodies preventing virus rebound . The paper is structured as follows . In section Methods we develop the general model of antibody responses to viral and the subviral particles . In section Analytical results we analyze it analytically using asymptotic analysis techniques and in section Numerical results we present numerical results and compare these to data of primary HBV infection in humans , and present alternative models . We conclude with a discussion .
Standard mathematical models of viral infections consider dynamics and relations among uninfected target cells ( ) , infected cells ( ) , and virus ( ) [25] , [26] , [27] , [28] , [29] . Briefly , these models assume that target cells become infected at a rate proportional to both the target cell concentration and the virus concentration , i . e . , at rate . Infected cells are thus produced at rate and die at rate ( that includes immune system mediated killing ) . Virus is produced by infected cells at rate and is cleared at rate . These models have been adapted to describe hepatitis B virus infection where the target population is liver cells ( hepatocytes ) [30] , [25] , [31] . The model accounts for the liver's ability to regenerate following injury [32] , [33] . This regeneration , accomplished by several cycles of hepatocyte mitosis , is described by a logistic term with carrying capacity and maximal proliferation rate [9] , [34] . Moreover , infected hepatocytes can be cured [10] and move back into the target population at rate [30] , [31] . The dynamics of the system are governed by the following equations ( 1 ) where for acute infection , and . An analysis of this system predicts two outcomes [35] , [36] , [37] . The infection dies out whenand the infection takes off and leads to chronic hepatitis when This simple system , which lacks any explicit immune response , does not explain transient infections , where the liver gets infected , i . e . , , but the infection is eventually cleared , presumably by the immune system . Clearance of HBV infection occurs in 90% of adults infected with HBV [1] . While the role of the cellular immune responses has been studied both theoretically and experimentally [9] , [34] , [38] , [39] , less is known about the dynamics of the humoral immune response to HBV [40] . In the following section we investigate antibody responses by modifying system ( 1 ) to account for humoral immunity following HBV infection . To include the antibody response , we generalize the model given by Eq . ( 1 ) by considering seven populations , corresponding to target cells ( ) , which are mostly or exclusively uninfected hepatocytes , productively infected cells ( ) , free virus ( ) , free subviral particles ( ) , free antibody ( ) , virus-antibody complexes ( ) , and subviral particle-antibody complexes ( ) . Since hepatocytes are in contact with the blood we assume , as above , that their infection can be described by a well-mixed system . Further investigation is needed to know whether spatial effects are important in HBV infection . For hepatitis C virus ( HCV ) infection , in which a much smaller fraction of cells become infected , spatial clustering of infected cells has recently been observed [41] . As in Eq . ( 1 ) , we assumed that target cells are maintained through homeostasis described by the logistic term with carrying capacity and maximum proliferation rate , and become infected at a rate proportional to both the target cell concentration and the virus concentration , i . e . , at rate . Infected cells are thus produced at rate and die at rate ( that includes immune system mediated killing ) . Upon infection , virus and subviral particles are produced at rates and , and cleared at rates and , respectively . We neglect the curing of infected cells by setting . With this simplification the basic reproductive number becomes . Previous papers [42] , [43] have presented detailed models of B lymphocyte proliferation and differentiation into plasmablast , antibody producing plasma cells and memory cells after they encounter antigen . For simplicity , we ignore the details of B-lymphocyte dynamics and differentiation into antibody producing cells and assume that free antibody ( ) , is produced at rate proportional to the antigen load , i . e , the viral and subviral concentrations , and is degraded at rate . Antibody is maintained after virus is cleared through antigen-independent homeostatic proliferation of memory B cells and long-lived plasma cells . In order to model this in a simple way , we add a logistic term to the antibody equation with maximum proliferation rate and carrying capacity . We show ( in section 4 . 3 ) that a model that explicitly includes B cell dynamics has behavior similar to this simpler model with antibody alone . Antigen elimination is facilitated by the formation of antigen-antibody complexes . We consider the reversible binding of free anti-HBsAg antibody ( ) , to both free virus ( ) , and subviral particles ( ) , described by the reaction scheme ( 2 ) where and are the complexes formed between antibody and the viral and subviral particles , respectively . , are binding rate constants , and , are disassociation rate constants for antibody reacting to viral and subviral particles , respectively . We consider that complexes and are degraded at a constant rates and . Antibodies can also bind infected cells budding virus . In this model we consider this to occur at a small rate and we neglect it . Based on the scheme ( 2 ) and the assumptions above we construct the following equations of virus-host interaction ( 3 ) where , , , , , and . The total concentration of viral DNA is described by ( 4 ) and the total concentration of anti-HBsAg antibody is given by ( 5 )
We developed a set of mathematical models of the antibody response to hepatitis B viral infection and tested whether viral clearance is possible in the presence of an excess number of subviral particles . Subviral particles are non-infectious but have HBsAg on their surface and thus bind anti-HBV antibody . If they bind enough anti-HBV antibody they have the potential to counter the antibody response . We used the models and data from seven acutely infected patients to determine important parameters that describe virus and antibody dynamics . Models of HIV have considered the antibody's effect on virus indirectly by modeling opsonization through enhanced viral clearance and neutralization through decreased virus infectivity [64] , [65] . Others considered in detail the interaction between virus and antibody that leads to complex formation [64] . Here , for the case of HBV , we modeled complex formation between antibody and both viral and subviral particles . This antibody model suggests that viral clearance is highly dependent on the characteristics of the antibody response: equilibrium antibody level , affinity , antigen-dependent and -independent antibody growth rates . The antibody model ( 7 ) predicts that , for the same antibody dynamics , virus clearance can occur for low SVP∶virus ratios but not for high SVP∶virus ratios . However , viral clearance can be achieved regardless of the SVP∶virus ratio if enough HBsAg-specific antibody is present at the time of infection or the antigen-dependent/-independent antibody growth rate is high enough . If we consider , as in clinical observations , that a healthy individual who produces an excess of -particles very likely clears the infection , than virus clearance occurs in the absence of pre-existent antibody when the antibody population's doubling time is faster than days ( Fig . 9 left panel ) . Moreover , the virus can still be cleared for slow antibody expansion if vaccine induced antibodies are present at the time of infection [66] . It is thought that following hepatitis B vaccination and boosting , patients with anti-HBsAg levels of mIU/ml ( mg/ml ) or higher are protected from infection [67] , [68] , [69] . In our study , if we assume an antibody population's doubling time of 16 . 6 days , the presence of HBsAg-specific antibody levels higher than mg/ml leads to virus clearance ( Fig . 9 right panel ) . We assumed that the antibody's carrying capacity is fixed at the maximum antibody levels observed after vaccination and boosting . Moreover , we assumed that the initial inoculum is low and fixed among all patients . Under these conditions we obtained virus clearance in the absence of pre-existent antibody in six out of seven patients . The predicted antibody levels needed for clearance , however , are higher than observed clinically in unvaccinated individuals , where free antibodies are usually only detected after HBV is nearly cleared . If the predicted free antibody levels are decreased substantially ( through decrease of activation parameters and or through decrease of the carrying capacity ) then pre-existent antibodies are required for protection ( Figs . S1 and S2 in Text S1 ) . The antibody model ( 7 ) predicts that more than of hepatocytes are infected at the peak of acute infection , lower than previously estimated [9] . These cells are replaced through homeostasis and the maximum liver loss at any one time ranges between . The length of time the continuous liver death and replacement occurs is dependent on the antibody dynamics , such as the antigen-dependent and -independent growth rates and , and the initial values , . Rapid liver cell turnover can lead to accumulation of mutations in the host genome that could result in genetic alterations , chromosomal rearrangements , activation of oncogenes , inactivation of tumor suppressor genes , and ultimately to hepatocellular carcinoma as seen in many patients with chronic hepatitis [70] . Model ( 7 ) assumed that one antibody is sufficient to neutralize a virion . We relaxed this assumption by developing a multivalent model ( 27 ) that accounted for multiple binding events . We tested whether the observed dynamics change when multiple antibodies bind a virion and consequently lower virus infectivity and/or enhance virus clearance . We found that such antiviral activity has an effect on the size ( but not the shape ) of free and bound virus during the second phase decay but not on viral peak ( Fig . 6 ) . Another antiviral response that we tested was the effect of increasing the ratio between the clearance of immune complexes and that of free virus . Originally we assumed that the immune complexes are cleared four times faster than the free virus , as in HIV [55] . In the virus clearance region , an increase in this ratio led to the decrease of free virus during the second phase decay ( Fig . S4 in Text S1 ) . An unresolved issue is whether the assay used to measure patients HBV levels ( the Amplicor HBV Monitor Test [57] ) measures only free virus , as we assumed , or both HBV DNA in free virus and immune complexes . In this regard , the patient HBV DNA was recovered from serum samples by a chemical denaturant method [56] , which has lower yield than techniques that use both chemical methods and proteolytic enzymes [71] . The latter should be more efficient at breaking apart virus-antibody complexes . We compared the estimates for the other parameters with those from our previous hepatitis B study that used the same patient data [9] . We found that the median infectivity rate is higher , and the viral production rate is smaller in the antibody model ( 7 ) . Our best estimates predict viral clearance in the first six patients and viral persistence in patient 7 . This is consistent with clinical results which report that patient 7 was immunosuppressed and developed chronic disease . Data fitting shows that he has the lowest infected cell clearance rates . The antibody model ( 7 ) shows that in order for the virus to be cleared this patient would need to have one or a combination of the following: ( 1 ) low subviral particles production together with a high pre-existing antibody level; ( 2 ) antibody of high affinity; ( 3 ) high antigen-dependent and/or -independent antibody growth; ( 4 ) increased clearance of virus-antibody complexes; ( 5 ) increased loss of infected cells . One important finding was that to obtain clearance of HBV , as observed in patients 1 to 6 and a majority of acutely infected adults , one needs high levels of antibody , higher than usually found in clinical observation . An alternative hypothesis for the effect of antibody is that they help clear the infection , once cellular immunity controls the initial burst of replication . The CD8-antibody model shows that cytotoxic effects lead to the initial viral control , and antibodies prevent re-infection . Previously , we had shown that a refractory/immune state of target cells could preserve the integrity of the liver while preventing re-infection [9] . Here we argue that antibodies could have a similar effect , with dynamics compatible with the observed levels and timing of antibodies in patients . Which of these effects is dominant , or if potentially both contribute to control of infection , needs further experimental studies . One limitation of model ( 30 ) is that , for patients 2 to 7 , killing of infected hepatocytes by CD8 T cells leads to more than liver loss . Therefore non-cytotoxic CD8 T cells effects leading to infected cells being cured and refractory to reinfection are needed to preserve liver integrity . In summary , we have developed a set of models of hepatitis B infection that give insight into the opposing roles of antibody and subviral particles in the resolution of acute HBV infection . In particular , we showed that even when the virus produces a large number of subviral particles , as a decoy against antibody protection , viral clearance can still be achieved when pre-existing immunity induced through vaccination or cross immunity leads to the presence of high antibody levels early in infection . However , in individuals with low initial antibody levels , as in most unvaccinated individuals and in individuals without prior exposure to HBV , antibodies could have more of a mop up function , clearing the infection and preventing viral resurgence after a cellular immune response controls the initial infection . | Hepatitis B vaccine induces life-long protection in vaccinated individuals . In the absence of vaccination , however , hepatitis B virus can cause both self-limiting and chronic disease . We investigate whether antibodies against hepatitis B play a role in virus clearance . We developed a mathematical model that describes the production of antibodies to both infectious virus and non-infectious subviral particles ( with hepatitis B surface proteins , but no nucleic acids ) and compared the model to patient data . We predict that high levels of antibodies , either pre-existing , as in vaccinated individuals , or through fast expansion , can control the infection and lead to viral clearance . However , when the antibody levels are more similar to those observed in a clinical context , cellular immune responses are needed to control the virus and antibodies act only in late stages to aid in viral clearance . | [
"Abstract",
"Introduction",
"Methods",
"Discussion"
]
| [
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"mathematics",
"applied",
"mathematics",
"biology",
"and",
"life",
"sciences",
"immunology",
"physical",
"sciences",
"immune",
"response"
]
| 2014 | Antibody Responses during Hepatitis B Viral Infection |
International melioidosis treatment guidelines recommend a minimum 10 to 14 days’ intravenous antibiotic therapy ( intensive phase ) , followed by 3 to 6 months’ oral therapy ( eradication phase ) . This approach is associated with rates of relapse , defined as recurrence following the eradication phase , that can exceed 5% . Rates of recrudescence , defined as recurrence during the eradication phase , have not previously been reported . In response to low eradication phase completion rates in Australia , a local guideline has evolved over the last ten years recommending a longer minimum intensive phase duration for many cases of melioidosis . This retrospective cohort study reviews antibiotic duration for the first episode of care for all patients diagnosed with melioidosis and surviving the intensive phase during a recent three year period in the tropical north of Australia’s Northern Territory; we also review adherence to the current local guideline and treatment outcomes . Of 215 first episodes of melioidosis surviving the intensive phase , the median ( interquartile range ) intensive phase duration was 26 ( 14-34 ) days . One hundred and eight ( 50 . 2% ) patients completed eradication therapy; 58 ( 27 . 0% ) patients took no eradication therapy . At 28 months’ follow-up , one ( 0 . 5% ) relapse and eleven ( 5 . 1% ) recrudescences had occurred . On exact logistic regression analysis , the only independent risk factors for recrudescence were self-discharge during the intensive phase ( odds ratio 6 . 2 [95% confidence interval 1 . 2-30 . 0] ) and septic shock ( odds ratio 5 . 3 [95% confidence interval 1 . 1-25 . 7] ) . Relapsed melioidosis is rare in patients who receive a minimum intensive phase duration specified by our guideline and extended according to clinical progress . Recrudescence rates may improve with reductions in rates of self-discharge . Given the low relapse rate despite a high rate of eradication therapy non-adherence , the duration and necessity of eradication therapy for different patients after guideline-concordant intensive therapy should be evaluated further .
Burkholderia pseudomallei , the cause of melioidosis , is a soil- and water-borne bacterium endemic to northern Australia and parts of south-east Asia [1] . It most commonly causes pneumonia , bacteremia without evident focus , deep-seated abscesses and skin infection but can affect almost any part of the body [2] . It causes more serious disease in patients with impaired innate immunity such as those with diabetes and has a high mortality rate ranging from 9% to 49% [3–5] . It is intrinsically resistant to most antibiotics and requires prolonged therapy for cure [5–8] . Current international guidelines suggest treatment with a minimum 10 to 14 days’ intravenous antibiotics ( intensive phase ) followed by 3 to 6 months’ oral antibiotics ( eradication phase ) [1 , 9 , 10] . Options for the intensive phase include ceftazidime or a carbapenem [10–13]; options for the eradication phase include trimethoprim-sulfamethoxazole ( TMP-SMX ) , doxycycline or amoxicillin-clavulanate [10 , 14–19] . Whilst there is provision in these international guidelines to extend the intensive phase to four weeks or greater in severe cases of melioidosis , the focus of the international guidelines is on switching to the eradication phase once the patient has been afebrile for 48 hours with negative blood cultures and an ability to take medication orally [9 , 10 , 17] . Despite treatment according to such guidelines , patients still have a high rate of relapse . In the Northern Territory ( NT ) of Australia , from 1989 to 2009 , 5 . 2% of 465 patients surviving the intensive phase have had molecularly confirmed relapse [4 , 20] . In Thailand , rates of relapse between 1986 and 2004 were at least 9 . 3% [5] . More recent data from a Thai randomized controlled trial demonstrated a relapse rate of somewhere between 1 . 1% and 6 . 4% [17] . Choice and duration of oral eradication therapy have been found to be the strongest risk factors for relapse [5] . A substantial proportion of our patients fail to complete the eradication phase and many live in remote communities making follow up difficult . In response to this , intensive phase therapy in our region has been progressively lengthened over the last 10 years . This is reflected in a local guideline developed at Royal Darwin Hospital ( RDH ) directing duration of therapy according to site of infection which in many cases extends intensive therapy far beyond defervescence with negative blood cultures and an ability to take oral antibiotics . In this study , we review antibiotic duration received by patients with melioidosis over a recent three year period , clinician and patient adherence to the local guideline and associated outcomes .
This study was a retrospective analysis of data from the ongoing Darwin Prospective Melioidosis Study [4 , 21] , a prospective cohort study . The study size was determined by the number of patients diagnosed with melioidosis during a recent period in and since which the guideline being evaluated has remained substantially unchanged . All patients with culture-confirmed melioidosis in the tropical Top End of the NT diagnosed between 1st October 2009 and 30th September 2012 who survived the intensive phase were eligible for inclusion in the study . Antibiotic therapy was managed by the Infectious Diseases Department at RDH . Exclusion criteria included cases that represented recrudescence or relapse of melioidosis first diagnosed prior to 1st October 2009 and cases with incomplete or inaccessible records . Antibiotic type and duration were reviewed for both the intensive and eradication phases along with demographic , clinical and laboratory data; these data had been recorded prospectively as part of the ongoing Darwin Prospective Melioidosis Study [4 , 21] . Intensive phase was defined as the period of time during which the patient received intravenous therapy directed against B . pseudomallei irrespective of clinician-intended duration . Eradication phase was defined as the period of time commencing at the end of the intensive phase and finishing at the guideline-recommended end date of eradication therapy irrespective of actual duration received . Intensive and eradication therapy referred to the intravenous and oral antibiotics directed against B . pseudomallei during the intensive and eradication phase respectively . Recrudescence and recurrence were defined as return of clinical illness during and after the eradication phase respectively with concomitant culture of B . pseudomallei from a clinical specimen . Recurrence was further defined as either relapse or reinfection when isolates from the initial and subsequent illness were identical or different respectively by multilocus sequence typing ( MLST ) [20] . Cure was defined as the absence of death during the eradication phase or recrudescence or relapse at the end of the follow-up period . Risk factors were defined as in the Darwin Prospective Melioidosis Study [4] . Antibiotic duration-determining focus ( ADDF ) was the focus requiring the longest minimum intensive phase duration according to the local guideline; if there were two or more such foci requiring the same minimum intensive phase duration , the ADDF was whichever of these appeared lowest on Table 1 , this being the focus generally considered most difficult to cure . Self-discharge was defined as voluntary cessation of inpatient status prior to completion of the clinician-planned intensive phase irrespective of guideline minimum duration . Non-adherence to eradication therapy was defined as cessation of eradication therapy prior to recrudescence or , if patients did not recrudesce , prior to the end of the eradication phase . The Darwin Melioidosis Guideline is summarized in Table 1 . First line intensive therapy was ceftazidime unless the patient was in the intensive care unit ( ICU ) or allergic or intolerant to ceftazidime in which case meropenem was used . If there were a collection ( including skin abscess or septic arthritis ) , bone or central nervous system ( CNS ) involvement , TMP-SMX , doxycycline or amoxicillin-clavulanate ( in order of preference ) was added early during the intensive phase for tissue penetration , usually orally . Subsequent oral eradication therapy used the same choice of these latter three antibiotics . Eradication phase was 90 days for each ADDF except osteomyelitis , CNS infection and arterial infection when 180 days was used . As noted in Table 1 , there was a small change to the guideline on 1st October 2010; clinician adherence was assessed against the final version of the guideline . Patients who survived the intensive phase were generally reviewed at monthly infectious diseases outpatient visits until completion of eradication therapy . Where patients failed to attend appointments , eradication therapy was assumed to have ceased at the last infectious diseases appointment attended or the last of any subsequent entries documenting therapy in the RDH medical record or NT-wide shared electronic health record . Follow-up data until 1st December 2014 were included; follow-up between the last clinic appointment and 1st December 2014 was performed retrospectively by reviewing hospital and community shared electronic health records . Melioidosis is a notifiable disease in the NT; data on melioidosis recurrence were based on Australia-wide laboratory notification of positive cultures to the NT public health unit . Data not normally distributed were expressed as median ± interquartile range ( IQR ) . Bivariate analysis of categorical and continuous variables was performed using the two-tailed Fisher exact test ( due to low expected cell values ) and the Wilcoxon rank-sum test respectively . Significant variables on bivariate analysis at p < 0 . 05 were assessed by multivariate analysis using exact logistic regression; exact methods were used due to the infrequency of recrudescence . Stepwise elimination of variables least significant on bivariate analysis was performed until all variables remaining in the model were statistically significant . Analyses were performed using Stata version 12 ( StataCorp , College Station , TX ) . This study was approved by the Human Research Ethics Committee of the NT Department of Health and the Menzies School of Health Research ( HREC 02/38 ) . As this was a retrospective observational study of a large dataset of a notifiable disease and data were analyzed anonymously , consent was not required .
Two hundred and fifty patients were diagnosed with melioidosis and managed by the RDH Infectious Diseases Department between 1st October 2009 and 30th September 2012 . Twenty-seven ( 10 . 8% ) of these died during the intensive phase , one was a relapse of a case diagnosed prior to 1st October 2009 and a further seven had incomplete antibiotic data due to missing files or partially interstate treatment; these patients were excluded leaving 215 patients in the study . All 215 patients were followed up as outlined in the methods and data analysis was performed on all patients except where stated . Baseline characteristics are shown in Table 2 . The median ( IQR ) age was 49 . 6 ( 39 . 0–60 . 5 ) years; 119 ( 55 . 3% ) patients were male and 15 ( 7 . 0% ) were under 18 years of age . The cohort had a high rate of comorbidity with 181 ( 84 . 2% ) patients having at least one recognized risk factor for melioidosis . In addition , melioidosis was of characteristic severity with 128 ( 59 . 5% ) patients bacteremic , 47 ( 21 . 9% ) requiring intensive care and 32 ( 14 . 9% ) developing septic shock . Intensive therapy duration is shown in Table 3 . The median ( IQR ) intensive phase duration for the 215 patients overall was 26 ( 14–34 ) days . Twenty ( 9 . 3% ) patients self-discharged during the intensive phase . Of 133 ( 61 . 9% ) patients who completed their intravenous therapy through the RDH Hospital in the Home program , the median ( IQR ) duration of infusor therapy was 14 ( 8–22 ) days . Median intensive phase duration according to ADDF is shown in Fig . 1 . When self-dischargers were excluded , patients with bacteremia with no focus or osteomyelitis as an ADDF tended to have a longer intensive phase duration than the guideline minimum duration . For patients with bacteremia with no focus , the most common reasons for prolonged therapy were immunosuppression ( 1 on cancer chemotherapy , 1 on high dose dexamethasone for cerebral metastases and 1 with systemic lupus erythematosus-related neutropenia ) and , in hindsight , incorrectly suspected other foci of infection ( 3 patients ) . Patients with osteomyelitis were more likely to have multifocal disease as depicted in Fig . 2 and tended to be slower to improve resulting in an extension of the intensive phase . Eradication phase duration was bimodal with one peak occurring at 90 days ( 93 [43 . 3%] patients ) and a second peak at 0 days ( 58 [27 . 0%] patients ) . Only 108 ( 50 . 2% ) patients completed the guideline-specified eradication phase . In total , 70 . 7% of eradication therapy days were with TMP-SMX; 29 . 3% were with doxycycline . No patients received amoxicillin-clavulanate . Adherence to the guideline-specified minimum intensive phase duration by clinicians according to ADDF is demonstrated in Fig . 3; self-discharging patients are excluded as it was not possible for clinicians to adhere in these cases . Of the remaining 195 patients , 43 ( 22 . 1% ) received less than the guideline-directed minimum intensive phase duration . In most cases of clinician non-adherence , the ADDF was skin abscess , pneumonia or deep-seated collection . Of the pneumonia cases , most had minor degrees of hilar/mediastinal lymphadenopathy and did not receive four weeks’ intensive therapy; of the deep-seated collection cases , most occurred in the first year of the study when it was common for patients to receive two rather than four weeks’ intensive therapy from last drainage . Despite this , 42/43 ( 97 . 7% ) clinician non-adherent cases had cure of their infection . Of the 215 patients , 197 ( 91 . 6% ) patients were cured . The median ( range ) duration of follow-up from the onset of the eradication phase for these cured patients was 45 . 9 ( 28 . 4–61 . 1 ) months . Six ( 2 . 8% ) patients died during the eradication phase . All six were palliated with eradication therapy ceased by the attending physicians prior to death . Five patients had incurable malignancy and one had advanced dementia . Two of the six patients had fever before dying but were not investigated due to their palliative status . A third patient , a 64 year old man with bacteremia with no focus , had been treated with four weeks’ ceftazidime followed by two weeks’ cotrimoxazole; this was then ceased due to a rash and was not replaced due to his palliative status in the context of squamous cell carcinoma of the lung with brain metastases , hypercalcemia and delirium requiring dexamethasone . He died 27 days after cessation of cotrimoxazole . Three days prior to dying he had a throat swab collected to investigate pneumonia; this grew B . pseudomallei 2 days after death . He was counted as a death during the eradication phase rather than a recrudescence as he was being actively palliated . The remainder of the six patients had no clinical evidence of infection prior to death . Eleven ( 5 . 1% ) patients recrudesced after their initial admission . The median ( IQR ) time to recrudescence from the end of the intensive phase was 24 ( 13–43 ) days; the median ( IQR ) time from the last day of taking eradication therapy was 10 ( 0–22 ) days with four ( 36% ) patients still taking their eradication therapy when they recrudesced . Of these four , three had an appropriate intensive phase duration according to the guideline but two of these recrudesced with infected foreign bodies that had not initially been evident clinically ( staghorn calculus and suspected vascular graft infection ) and the third had bronchiectasis with persistent sputum positivity . Only one ( 0 . 5% ) relapse occurred . This was a 34 year old man with pneumonia , mediastinal lymphadenopathy and an undrained liver abscess who self-discharged after 15 days’ intensive therapy and took no eradication therapy . He relapsed 15 months later with severe sepsis due to worsening pulmonary abscesses and mediastinal lymphadenopathy but had a stable-appearing liver abscess . One ( 0 . 5% ) patient had reinfection . He was considered for the purpose of analysis to have had cure of his original infection . There were no patients with clinically-suspected but culture-unproven recrudescence or recurrence . Bivariate analysis was performed to assess for significant predictors of recrudescence . Patients who died during the eradication phase were excluded due to a shorter exposure to the possibility of recrudescence and a likely bias toward not diagnosing recrudescence due to the palliative nature of these patients . This left 209 patients in the analysis . Results of analysis of categorical variables are shown in Table 4 . The median ( IQR ) age for patients who did and did not recrudesce was 46 . 3 ( 41 . 2 to 55 . 6 ) and 49 . 6 ( 38 . 8–59 . 9 ) years respectively ( p = 0 . 47 ) . The only variables significantly predicting recrudescence were diabetes , having osteomyelitis as an ADDF , admission to ICU , septic shock and self-discharge . Multivariate analysis showed that only two of these variables were statistically significant independent predictors of recrudescence; these were self-discharge ( odds ratio 6 . 2 [95% confidence interval 1 . 2–30 . 0 , p < 0 . 05] ) and septic shock ( odds ratio 5 . 3 [95% confidence interval 1 . 1–25 . 7 , p < 0 . 05] ) . A subgroup analysis was performed on the 58 patients who took no eradication therapy; 52 ( 89 . 7% ) were cured , 2 ( 3 . 4% ) died during the eradication phase , 3 ( 5 . 2% ) recrudesced and 1 ( 1 . 7% ) relapsed .
We have developed a guideline for duration of intensive phase therapy for melioidosis that we think is responsible for the very low rate of relapse now seen in the Top End of the NT . Its immediate applicability in some developing regions is uncertain due to funding and resource issues . However , given the excellent cure rates with our intensive phase guideline despite poor adherence to subsequent eradication therapy , we believe that further research evaluating the duration and necessity of the eradication phase for different ADDFs is now warranted . | Melioidosis is an infection caused by the soil bacterium Burkholderia pseudomallei; patients usually present with pneumonia , blood-stream infection and/or skin or internal organ abscesses . Melioidosis occurs most commonly in northern Australia and parts of Southeast Asia . It has a high mortality rate and , with standard treatment , a relapse rate greater than 5%; patients who relapse often represent severely unwell . Treatment comprises an intensive ( intravenous antibiotic ) phase , followed by a prolonged eradication ( oral antibiotic ) phase . Previous studies have found that the intensive phase is important to prevent mortality , and the eradication phase is important to prevent relapse . However , these studies have not been designed to detect an effect of intensive therapy on relapse rate . We know that adherence to eradication therapy is poor , and many of our patients live remotely making follow-up difficult . In order to address this , we have developed a new treatment guideline which stipulates a longer intensive phase for most patients . We show that adherence to this guideline is associated with very low relapse rates despite poor adherence to eradication therapy . It is possible that for many patients the eradication phase could be shortened or avoided when this intensive phase guideline is followed; this requires further research . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
]
| []
| 2015 | Intravenous Therapy Duration and Outcomes in Melioidosis: A New Treatment Paradigm |
Noroviruses are major pathogens associated with acute gastroenteritis worldwide . Their RNA genomes are diverse , with two major genogroups ( GI and GII ) comprised of at least 28 genotypes associated with human disease . To elucidate mechanisms underlying norovirus diversity and evolution , we used a large-scale genomics approach to analyze human norovirus sequences . Comparison of over 2000 nearly full-length ORF2 sequences representing most of the known GI and GII genotypes infecting humans showed a limited number ( ≤5 ) of distinct intra-genotypic variants within each genotype , with the exception of GII . 4 . The non-GII . 4 genotypes were comprised of one or more intra-genotypic variants , with each variant containing strains that differed by only a few residues over several decades ( remaining “static” ) and that have co-circulated with no clear epidemiologic pattern . In contrast , the GII . 4 genotype presented the largest number of variants ( >10 ) that have evolved over time with a clear pattern of periodic variant replacement . To expand our understanding of these two patterns of diversification ( “static” versus “evolving” ) , we analyzed using NGS the nearly full-length norovirus genome in healthy individuals infected with GII . 4 , GII . 6 or GII . 17 viruses in different outbreak settings . The GII . 4 viruses accumulated mutations rapidly within and between hosts , while the GII . 6 and GII . 17 viruses remained relatively stable , consistent with their diversification patterns . Further analysis of genetic relationships and natural history patterns identified groupings of certain genotypes into larger related clusters designated here as “immunotypes” . We propose that “immunotypes” and their evolutionary patterns influence the prevalence of a particular norovirus genotype in the human population .
RNA viruses evolve quickly , with mutation rates that vary between 10−3–10−4 nucleotide ( nt ) substitutions/year; up to 1000 times higher when compared with most DNA viruses [1] . This high mutation rate is attributed largely to the inability of their RNA polymerases to correct errors introduced during replication . In addition to nt substitutions , RNA viruses generate diversity by recombination or rearrangements of their genome [2–4] . This diversity is considered required for the virus to ( i ) escape the immune surveillance of the host [5–7] , ( ii ) reach different organs of the host and/or change tropism , and ( iii ) induce the pathological effects that could lead to efficient transmission to a new host [8] . Although extreme diversity might be advantageous for a virus to exploit different niches , viruses can also retain their phenotypic characteristics for decades ( e . g . , dengue virus and respiratory syncytial virus ) [5 , 9] . Noroviruses are a major cause of acute gastroenteritis worldwide , mainly associated with outbreaks occurring in closed settings , such as hospitals , nursing homes , schools , cruise ships , and military facilities . In countries where rotavirus vaccination has been successfully introduced , norovirus has become the major cause of gastroenteritis in children [10–13] . It is estimated that noroviruses cause between 70 , 000 to 200 , 000 deaths per year worldwide , with the majority in children from developing countries [14 , 15] . Although symptoms characteristically resolve within 72 hours in healthy immunocompetent individuals , norovirus genomic RNA can be detected in stool for up to 2 months [16–19] , and clearance of virus seems to be associated with a specific immune response in the mucosa [18] . In immunocompromised patients , symptoms and virus can persist over years , complicating the management of their underlying disease [20] . Noroviruses are considered fast-evolving viruses [21–24] , and present an extensive diversity that is driven by acquisition of point mutations and recombination . The genome consists of a single-stranded positive-sense RNA molecule of ~7 . 5kb that is organized into three open reading frames ( ORFs ) . ORF1 encodes a polyprotein that is co-translationally cleaved into six proteins required for replication , while ORF2 encodes the major capsid protein ( VP1 ) , and ORF3 , a minor capsid protein ( VP2 ) [25] . Expression of recombinant VP1 yields virus-like particles ( VLPs ) that mimic the native virion , which have been important research tools in the absence of tractable cell culture systems and animal models [26–30] . Based on sequence differences of the VP1 protein , noroviruses have been classified into seven genogroups ( GI-GVII ) and over 30 genotypes [25 , 31] . Despite this extensive diversity , a single genotype ( GII . 4 ) has been shown to be the most prevalent in humans worldwide [31–33] . Since the mid-1990s , six global epidemics have been documented and each has been associated with the emergence of a new GII . 4 variant . The first was characterized by an increased number of norovirus outbreaks worldwide , and associated with the US95_1996 virus . The second epidemic started in 2002 and coincided with the replacement of the US95_1996 virus by the Farmington_Hills_2002 virus . The third epidemic was caused by the Hunter_2004 virus , which was rapidly replaced by two new pandemic strains , namely Den_Haag_2006b and Yerseke_2006a . During 2009 , a new strain emerged ( New_Orleans_2009 ) that co-circulated with the 2006 variants for almost three years until the current predominant variant emerged in 2012 ( Sydney_2012 ) [25 , 31 , 33] . Interestingly , this epidemiological pattern has not been reported for any other norovirus genotype , until recently . A novel GII . 17 variant has emerged , potentially displacing an older GII . 17 , causing large outbreaks in different countries from Asia [22 , 34–36] . Although GII . 4 is overall the most prevalent genotype in the human population , multiple norovirus genotypes co-circulate in children with low to high incidence . Genotypes GII . 3 , GII . 6 and GII . 2 ( in addition to GII . 4 ) have consistently been linked to infection in children under 5 years of age [17 , 21 , 37 , 38] . Initial challenge studies in human volunteers suggested a lack of protective responses between strains from the two major genogroups ( GI and GII ) , as cross-challenge between Norwalk virus ( the GI . 1 prototype strain ) and Hawaii virus ( the GII . 1 prototype strain ) did not induce protection . In addition , duration of immunity might be short ( less than 6 months ) , as individuals re-challenged with the same virus became ill during the second exposure [39] . It has been noted that the high titer of challenge virus administered in the early volunteer studies might not reflect that during natural exposure [40 , 41] and recent studies have focused on the natural history of noroviruses . Based on epidemiological data , Simmons et al . modeled that norovirus genotype-specific immunity could last up to 9 years [42] , which would enhance the duration of vaccine-induced immunity . The diversity of genotypes has also been addressed . Children can be re-infected multiple times during the first 5 years of life [17 , 18 , 38] , with the majority of re-infections occurring with different norovirus genotypes . These data suggest that genotypes may represent distinct serotypes , which would complicate vaccine design . In this study , we integrated large-scale genomics analysis with natural history data to investigate mechanisms involved in the diversification and evolution of norovirus genotypes and their variants . Most norovirus intra-genotypic variants displayed a striking genetic stability over long periods of time , with GII . 4 as the notable exception . We detected patterns of re-infection and susceptibility consistent with genetic and antigenic clustering of certain genotypes and propose that these relationships may be relevant in the design of norovirus vaccines .
To investigate the diversity and evolutionary differences of the distinct norovirus genotypes , more than 2000 sequences of the gene ( ORF2 ) encoding the VP1 were retrieved from GenBank , with 101 and 1909 genes from GI strains and GII strains , respectively ( Table 1 ) . Individual sequences were vetted with an online norovirus typing tool that follows a widely-used universal classification and nomenclature system for norovirus genotypes [25] , and an effort was made to verify the date of occurrence with the supporting documentation . Genotypes with 10 or more complete ( or nearly complete ) ORF2 sequences were selected for further phylogenetic analysis at both the nt and amino acid ( aa ) level , and included 16 out of the 31 current GI and GII genotypes ( Table 1 ) . We defined an intra-genotypic variant as a group of strains ( ≥ 2 ) that clustered together in the phylogenetic tree and that showed <5% difference in their nt or aa sequences , but ≥5% difference compared to other strains . Most genotypes segregated into 1 to 5 phylogenetic variants when nt sequences were analyzed ( Table 1 ) , with the exception of genotype GII . 4 that displayed at least 10 different variants ( Table 1 and Fig 1A ) . The number of variants was lower in seven genotypes ( GI . 1 , GI . 4 , GI . 6 , GII . 2 , GII . 12 , GII . 13 and GII . 14 ) when aa sequences were used for tree reconstruction and distance analyses ( Table 1 ) . To link the different intra-genotypic variants to phenotypic characteristics in the VP1 , we focused on the analysis of aa sequences . Interestingly , non-GII . 4 genotypes presented variants with strains that have been detected many years apart ( mean: 24 . 9 years , standard deviation: 12 . 97 years ) while having only a few differences in their aa sequence ( Table 1 ) . An example is the GII . 6 genotype , with each of the three variants ( A-C ) containing strains that differed by only a few aa residues ( ≤1 . 2% ) but that were detected up to 41 years apart ( Fig 2A ) . In contrast , the GII . 4 genotype was comprised of variants that were present in the human population from 3 to 8 years ( mean: 5 . 3 years , Fig 1A ) . We developed an algorithm to illustrate visually the relationship between amino acid diversity and time among strains from a given genotype . The algorithm generated a heat map in which each square represents the number of strains with such aa difference in their VP1 and plotted against the timespan of detection . Analysis of 16 genotypes with sufficient data in GenBank revealed two distinct patterns of variant evolution; one in which the number of aa differences accumulated continually over time ( Fig 1B ) , and one in which the number of aa differences remained relatively constant over time , regardless of the timespan between strains ( Fig 2B ) . The first pattern ( Fig 1B ) was related to a constantly changing “evolving genotype” represented by the GII . 4 , while the second pattern ( Fig 2B ) was a highly conserved or “static genotype” represented by the 15 other genotypes with sufficient data for analysis ( Table 1 ) . The static genotypes resolved readily into distinct intra-genotypic variants , with one exception . The diversity plot of the GII . 12 genotype strains displayed a subtle accumulation of differences over a time period of approximately 20 years; however , those differences were not associated with different intra-genotypic variants ( S1 Fig ) . A larger number of sequences over a longer period will be helpful for defining variant diversity in GII . 12 and other static genotypes . During 2014–2015 , a sharp increase in the number of gastroenteritis outbreaks was reported in Asia [22 , 34 , 43] that were associated with the emergence of a new variant of the genotype GII . 17 , i . e . variant C or Kawasaki_2014 [35 , 43] . Our phylogenetic analysis of this genotype revealed four distinct variants , with one of the variants having strains that spanned over 37 years with a low level of sequence diversity consistent with a static pattern ( S2 Fig ) . In contrast , the emerging strains showed multiple substitutions at the nt and aa sequence level [22 , 35][44] , which led to diversification into two separate phylogenetic variants ( C and D ) . These differences appeared to accumulate over time , with predominant strains that circulated during the 2014–2015 season ( variant D ) differing by 5 . 4±1 . 1% from those in the 2013–2014 season ( variant C ) in their aa sequence . To gain insight into the evolutionary differences noted at the aa level , the nt rate of evolution and nonsynonymous substitutions ( dN ) /synonymous substitutions ( dS ) ratio were calculated for each of the genotypes included in this analysis . The nt rates of evolution were similar among all the norovirus genotypes ( range: 5 . 40x10-3–2 . 23x10-4 nt substitutions/site/year ) . However , differences were seen in the dN/dS ratios , with genotypes GII . 4 and GII . 17 presenting slightly higher values than any other norovirus genotype ( Table 1 ) . Note that in five different genotypes ( GI . 3 , GI . 4 , GII . 4 , GII . 13 , and GII . 17 ) , the dN/dS ratio in the P2 encoding region was at least two times higher than the complete ORF2 ( encoding VP1 ) dN/dS ratio . The dN/dS ratio has been used to inform the evolutionary pressures on a gene: a dN/dS > 1 ( higher number of non-synonymous mutations ) indicates positive evolution as the phenotype is changing due to pressures of the environment ( e . g . immune responses ) , while a dN/dS < 1 indicates a purifying selection ( also known as negative selection ) , where the new phenotype is mostly deleterious and eliminated from the population [45–47] . Despite the small differences in the dN/dS ratios , purifying selection ( dN/dS < 1 ) is strongly acting at the VP1 protein level in all norovirus genotypes , including GII . 4 . We next examined whether the different patterns of diversification observed in ORF2 extended to other regions of the genome . We analyzed by NGS full-length norovirus genomes from immunocompetent individuals infected in different settings . The first set of samples was from a child who was consecutively infected with three different genotypes ( GII . 4 , GII . 6 and GII . 17 ) over a 3-year period [18 , 35] . Although in each episode the child resolved the symptoms within ~ 72 hours , viral RNA was detected in stools for weeks after onset of symptoms . The full-length genome sequences were compared within the first 3 weeks for GII . 4 and GII . 6 viruses and the first 2 weeks for GII . 17 viruses . A total of up to 78 , 000 reads/site ( mean: 14688 , standard deviation: 5675 ) were obtained for each sample by NGS ( S1 Table ) . All consensus sequences were identical to the reference ( day 1 [d1] ) ; however , mutations ranging from 5 to 50% of the total reads ( S1 Table and Fig 3 ) were found at later time points for each virus . In GII . 4 viruses , 12 nt mutations arose in the subpopulations of the sample collected at d14 , with nine of them being non-synonymous mutations . Three aa mutations were located in the P domain , the capsid surface-exposed region of the VP1 protein , with two present in antigenic sites A ( E368A ) and E ( S412N ) [48] ( S3 Fig ) . By d21 , the GII . 4 virus had acquired 20 nt mutations in its subpopulations , with most of them ( 19/20 ) new mutations as compared to the d14 sequence . Nine out of the 20 nt mutations changed the aa sequence , with 4 of them mapping to the P domain , near or at antigenic sites ( S3 Fig ) . In contrast , GII . 6 and GII . 17 viruses presented only four and two substitutions in their subpopulations at d21 and d14 , respectively , with no evidence of the accumulation of mutations over time ( S2 Table ) . Although a large amount of norovirus is shed in stool , the infectious dose in natural transmission is likely low [41] . Thus , during inter-host transmission events noroviruses may undergo an initial reduction in the number of replicating viruses , creating a bottleneck effect . To compare inter-host evolution in individuals involved in outbreaks , we analyzed samples from outbreaks that occurred in the state of Maryland where the causative agents were identified as GII . 6 ( a hospital outbreak in 1971 ) [49] or GII . 4 noroviruses ( nursing home outbreaks in the 1987–1988 winter season ) [32] . The samples and their dates of collection are indicated in Fig 4 . Comparison of the NGS sequences with the outbreak consensus sequence revealed only a few substitutions , ≤ 5 nt and ≤ 2 aa , among samples from the same outbreak ( Fig 4A and Fig 4B ) . However , when the consensus sequences from the GII . 4 outbreaks were compared , a progressive accumulation of mutations ( up to 86 nt and 16 aa ) were detected in a period of three months , with no aa substitutions detected in the ORF2 ( Fig 4B ) . To confirm these observations , we compared 151 genomes from GII . 4 viruses ( variant Den Haag ) detected during three epidemic seasons in Japan [50 , 51] , and showed the accumulation of nt and aa substitutions over time ( S4 Fig ) . In both sets of samples , the ORF2 ( encoding VP1 ) acquired fewer amino acid substitutions as compared with ORF1 and ORF3 , and thus maintained the VP1 phenotype for the GII . 4 variant circulating in that given season . The data from full-length genome analyses were consistent with those from analyses of ORF2 in the GenBank database: different patterns of evolution exist among the norovirus genotypes in an acute outbreak setting . To reconcile our observations on the different mechanisms of diversification and data on re-infection and epidemiology of noroviruses , we investigated whether additional relationships might exist among the genotypes from the two major genogroups . Our phylogenetic tree constructed with representative strains from each genotype ( strains from each lineage described here were included ) showed clustering among certain genotypes ( e . g . GI . 3 , GI . 7 , GI . 8 and GI . 9 ) , while others appeared as single genotypes ( e . g . GI . 1 , GII . 3 , GII . 6; Fig 5A , S3 Table ) . The genotypic clustering was reproducible with a second phylogenetic methodology ( S5 Fig ) . We designated each of the separate branches as groups A-L ( Fig 5A ) , and the deduced aa sequences showed an approximate cut-off value of ≥20% aa differences between groups ( Fig 5B ) . In a review of data from research groups that have documented norovirus re-infections and determined the genotype for each infection [17 , 18 , 38 , 52–54] , we observed that the pattern of re-infection might be consistent with the new grouping system as a predictor of antigenically-distinct strains . To test the hypothesis that these groups , provisionally designated here as “immunotypes , ” might play a role in norovirus immunity , we developed a matrix that recorded the data from each of the consecutive re-infection cases with documented norovirus genotyping [17 , 18 , 22 , 38 , 52–54] . For example , a child consecutively infected with a GII . 4 ( “immunotype” G ) , GII . 6 ( “immunotype” H ) , and GII . 17 ( “immunotype” J ) norovirus would count as one individual for the cell of the matrix that compares immunotype G and H , and as one individual for the cell that compares immunotype H and J . A matrix was constructed using re-infection data available from 116 children and 2 adults ( Fig 5C , S6 Fig ) . Overall , the majority of re-infections occurred with viruses from different immunotypes , with re-infection rare from strains within an immunotype . A notable exception was immunotype G , which is comprised of genotypes GII . 4 and GII . 20 . Re-infection of eight children ( as shown by the 8 individuals in the black cell ) was documented to have occurred with different variants of GII . 4 viruses [17 , 38] .
Viruses are genetically and structurally diverse . Depending on their genome and/or replication strategies , viruses can present different rates of evolution ( range: 10−2–10−9 nt substitutions/site/year ) [3 , 47 , 55] . As with many other RNA viruses , noroviruses have been regarded as rapidly evolving viruses [22 , 48 , 56] . The overall rate of evolution for the norovirus genotypes included in this study ranged from 5 . 40x10-3–2 . 23x10-4 nt substitutions/site/year for the VP1 encoding region , which were similar to those described previously for norovirus GII . 4 , GII . 3 , GII . 6 , and GI . 1-GI . 6 [21 , 57–61] , and within the range for positive-strand RNA viruses [3] . Despite this high nt mutation rate , the number of non-synonymous substitutions were on average ~18 times lower than the synonymous substitution ( dN/dS average: 0 . 06 ) , suggesting that purifying selection ( dN/dS <1 ) acts strongly in the VP1 protein . Similar observations have been made for other RNA viruses , where the rate of evolution reached up to 10−2 nt substitutions/site/year ( depending on the region of the genome used for analyses ) but was mostly dominated by high synonymous substitution rates [46 , 55 , 62] . In noroviruses , positive selection has been reported for certain codons of the VP1 for GII . 4 , GII . 3 , GII . 6 and GII . 17 viruses [21 , 22 , 57 , 58 , 63] , and codon changes in the antigenic sites of GII . 4 viruses ( which are located in loops of the P2 domain ) have correlated with the emergence of new variants [24 , 27 , 48] . Taken together , our findings suggest that the capsid protein of all noroviruses evolve with strong structural constraints , with only a limited number of codons that can evolve and , perhaps confer adaptive advantages to infect human hosts . Epidemiological studies coupled with sequence data from field isolates have indicated that the most predominant norovirus genotype , GII . 4 , is evolving similarly to influenza H3N2 viruses; i . e . with a temporal replacement of predominant variants that is driven by the immune response of the host [27 , 48 , 64] . By exploring the intra-genotypic diversity from representative human norovirus genotypes we verified that GII . 4 noroviruses produce the largest number of intra-genotypic variants , and that these variants last ( on average ) ~5 years in the human population . In contrast , non-GII . 4 noroviruses sustain a low number of intra-genotypic variants with a limited number of aa differences among strains within that given variant; even if decades apart in occurrence . Interestingly , different variants from a given genotype can often be co-circulating within the same year and geographical location causing gastroenteritis [37 , 44 , 65 , 66] . The GII . 4 viruses , and to a lesser extent one variant of the GII . 17 viruses , acquired aa substitutions over time that created phenotypically different variants . In contrast , all other genotypes retained similar sequences within variants that might have arisen early in the origin of that genotype and that persisted over time . This led us to discriminate two different patterns of evolution in norovirus: evolving and static . Evolving viruses continually accumulate mutations in their genome over time , and static viruses do not . The concept of evolving versus static norovirus genotypes may be helpful in understanding the spread of pandemic strains . The recent emergence of GII . 17 viruses resulted in the rapid replacement of one variant ( variant C ) with another ( variant D ) [22 , 44] . This pattern of very rapid replacement , occurring within two consecutive seasons , in the emerging GII . 17 viruses is notably different from that of GII . 4 viruses , in which each emerging GII . 4 variant is replaced every 3 to 8 years . Thus , since the GII . 17 genotype presents other variants shown to be “static , ” the recent global spread of the GII . 17 genotype might be the moment when a new genotypic variant ( variant C ) emerged and is quickly adapting to reach maximum fitness in the human host ( variant D ) to become static . Since the emergence of this GII . 17 strain has only recently occurred and most of the available GII . 17 sequences ( 136/143 ) correspond to these two variants , more information on pre-2013 strains and the future epidemiological behavior of the GII . 17 strains will be helpful in establishing the evolutionary pattern of this genotype . Because recombination has been suggested to play an important role on the emergence of many GII . 4 variants [67] , and the emerging GII . 17 strains presented a novel polymerase ( encoded by ORF1 ) [22 , 34 , 35 , 44] , further studies should be conducted on the role of recombination in norovirus VP1 diversification into variants . To determine the role of intra-host evolution at the genomic level , we developed a method to generate and analyze full-length norovirus genomes with NGS technologies and bioinformatics . The strategy of amplification was similar to that published by Eden et al . [67] for GII . 4 viruses , and our method was robust for a number of GII noroviruses ( GII . 1 , GII . 2 , GII . 3 , GII . 4 , GII . 6 , GII . 12 , and GII . 17 ) , and from samples stored for over 40 years [35] . Several groups have explored the intra-host diversity of noroviruses by NGS using partial regions of the genome [23 , 68]; however , our approach extended these findings by allowing high-resolution analysis at every nt position in the coding sequence of the genome . We first examined the intra-host evolution of GII . 4 , GII . 6 and GII . 17 noroviruses within a single patient , and observed that only the GII . 4 viruses presented a gradual increase in the number of mutations , which in some cases resulted in aa substitutions in areas regarded as important antigenic sites . The limited intra-host diversity found during the shedding phase of an infection in immunocompetent individuals contrasts with the vast diversity of viruses found in immunocompromised patients [68] . Due to the diversity found in immunocompromised patients and prolonged shedding , it was suggested that they might be a source of new GII . 4 variants to the human population [69] . Noroviruses are highly transmissible; however , there is little evidence that norovirus can be efficiently transmitted during the chronic phase of the infection [19] . A more likely source for new GII . 4 variants might be immunocompetent individuals , where we show that mutations can arise during inter-host transmission events , and accumulate during the intra-variant period . Although noroviruses belonging to the “static” genotypes can also accumulate mutations during inter-host transmission events , those mutations would likely be eliminated from the viral population by purifying selection . Viruses that better tolerate the introduction of mutations are regarded as genetically robust , and this robustness has been shown to be beneficial for virus survival and prevalence [70] . Overall our data suggest that GII . 4 noroviruses are genetically robust . In contrast , noroviruses with “static” genotypes may be genetically fragile , which limits their antigenic diversity and prevalence . How do “static” genotypes prevail in the human population , in the face of limited antigenic diversity within the genotype ? To address this question , genotypes were grouped together based on phylogenetic clustering and aa differences in their capsid proteins . These groups , or “immunotypes , ” were applied to the interpretation of epidemiological observations . When examining data from a birth cohort study , or reports where children and adults were followed for years to study norovirus re-infections , genotypes belonging to the same immunotype generally did not re-infect these individuals . Thus , most of these individuals were re-infected with a varying series of genotypes ( predominantly containing combinations of GII . 4 , GII . 6 , GII . 3 , GII . 17 or GII . 2 ) , but all of them belonging to different immunotypes as defined in Fig 5C . The exception to this was the GII . 4 strains in immunotype G , in which a few re-infections were observed , albeit with different GII . 4 variants . Based on these data , we propose a model for norovirus re-infection in which naïve children are constantly exposed and infected with strains from each of the different immunotypes until a broad immunity develops . In contrast , older individuals ( i . e . older children and adults ) are more likely to become ill from evolving genotypes , as they have already acquired immunity against a number of static genotypes ( Fig 6 ) . This model not only explains the differences in the genotype distribution often seen when comparing children and adult populations [17 , 37 , 38] , but also suggests that immunity against norovirus may be longer than initially suggested [39 , 42] . For decades understanding of norovirus immunity was based on human volunteer challenge studies and animal models or in vitro surrogates of neutralization tests [27 , 39 , 71 , 72] . Initial cross-challenge studies , conducted in the 1970s using the prototype GI . 1 Norwalk virus and GII . 1 Hawaii virus , showed a lack of protection between these two genogroups [39] . Further epidemiological data and in vitro assays , such as antibody blockage of carbohydrate binding to VLPs , suggested a role for immunity against the different intra-genotypic variants of GII . 4 [27 , 33 , 58] . Norovirus vaccines are currently based on the premise to include at least two major antigens for noroviruses representing GI and GII [29 , 30 , 71 , 73 , 74] . However , recent data indicating that certain genotype-specific immune responses were unable to confer natural protection against disease raised concerns that a prohibitive number of components ( almost 30 ) might be needed in a norovirus vaccine [17 , 18 , 35 , 38] . Although additional studies will be needed to confirm the existence of shared antigenic groups among the norovirus genotypes , preferably by neutralization assays or animal models , our analysis provides a new perspective on the genetic and antigenic diversity of noroviruses that could lead to the identification of cross-protective strains and inform vaccine design .
Stool specimens from the child were obtained with the written informed consent of the parent , and enrollment in National Institutes of Health ( NIH ) clinical study NCT01306084 . Archival stool samples stored in the Laboratory of Infectious Diseases Calicivirus Repository were waived as exempt from IRB review by the NIH Office of Human Subjects Research and Protection ( OHSRP 11833 ) . Epidemiological information relating to the sample collection has been published elsewhere [18 , 32 , 35 , 49] . The full-length ( or nearly full-length ) ORF2 sequences ( encoding for VP1 ) from each of the 31 genotypes described for GI and GII were retrieved from GenBank ( accessed on March 2015 ) for analyses . Alignments were performed with Clustal W as implemented in MEGA v6 [75] . Sequences from each genotype were aligned separately to minimize the presence of insertions or deletions ( indels ) , which can arise when different genotypes are compared . Phylogenetic trees were constructed using Kimura 2-parameter as method of nt substitution and Neighbor-Joining as algorithm of reconstruction as implemented in MEGA v6 with default settings . Phylogenetic trees that used aa sequences were reconstructed using a Poisson method of aa substitution . Bootstrap analyses were used to support the clustering of the variants . Information on the strains used for the phylogenetic analyses is provided in S4 Table . Evolutionary rates ( nt substitutions/site/year ) for each genotype were estimated using the ORF2 sequences and the Bayesian Markov Chain Monte Carlo ( MCMC ) approach as implemented in the BEAST package [76] . For each set of data the General Time Reversible ( GTR ) model with gamma rate distribution and invariable sites parameter was used and the MCMC was run for a sufficient number of generations to reach convergence of all parameters . All evolutionary rates were calculated using strict clock model and coalescent constant size tree prior , except for genotypes GI . 4 , GI . 6 , GII . 14 and GII . 16 , which reached convergence using Bayesian Skyline and random local clocks . Selection pressures acting in the VP1 sequences were investigated by estimating the mean rate of nonsynonymous substitutions ( dN ) and synonymous substitutions ( dN ) and the dN/dS ratio as implemented in MEGA v6 . The nearly full-length genome sequences from 151 GII . 4 viruses detected in Japan during 2006–2009 [50 , 51] were downloaded from GenBank and analyzed using MEGA v6 and Prism software ( GraphPad Prism version 7 ) . To visualize the aa substitutions within each genotype , a Python script ( available upon request ) was developed to calculate the number of aa differences and the isolation year differences between two individual strains . Isolation years were extracted from strain descriptions . The difference values were added into a matrix where the y-axis represents the isolation year differences and the x-axis the amino acid differences . Note that some cells will present more than one comparison , since strain pairs presenting the same number of aa differences and the same year difference , despite the years detected , will be included in the same cell . Heat map plots were calculated for each genotype using GraphPad Prism version 7 ( GraphPad Software , La Jolla California USA ) , with the values representing the number of strains compared . A platform was developed to analyze the plasticity of norovirus genotypes at the full genome level . Briefly , viral RNA was extracted from 10% ( w/v ) stool suspensions using the MagMax Viral RNA Isolation Kit ( Ambion , California , USA ) following manufacturer’s recommendations . Complementary DNA was synthesized from the viral RNA using the Tx30SXN primer ( GACTAGTTCTAGATCGCGAGCGGCCGCCCTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT [77] ) at 5μM final concentration , and the Maxima H Minus First Strand cDNA Synthesis Kit ( Thermo Fisher Scientific , California , USA ) following manufacturer’s recommendations except that only 0 . 1 μL of Enzyme Mix was used per reaction . Amplification of the full-length genome was performed using 5 μl of the RT reaction , a set of primers that target the conserved regions of the 5’- and 3’-end of GII noroviruses ( GII1-35: GTGAATGAAGATGGCGTCTAACGACGCTTCCGCTG , and Tx30SXN ) , and the SequalPrep Long PCR Kit ( Invitrogen , California , USA ) following manufacturer’s recommendations . Amplicons were excised from an agarose gel and purified with the QIAquick Gel Extraction Kit ( Qiagen , California , USA ) . Ion Torrent libraries were prepared by using 300–500 ng of full-length genome PCR amplicons following standard Ion Torrent library prep protocol . DNA was fragmented followed by the introduction of ligation barcode adapters . Adapted-ligated libraries were amplified using 13 PCR cycles , and size selected from agarose gels . Final libraries were quantified by Qubit ( Invitrogen , California , USA ) , Bioanalyzer ( Agilent ) , and qPCR . Libraries were normalized to 1nM , pooled at an equal molar ratio , and loaded onto a 318 v2 Chip in an Ion OneTouch2 machine . The sample from Ion OneTouch2 was transferred to an Ion OneTouch ES and then to an Ion PGM for sequencing with a 400bp kit ( Life Technologies , California , USA ) . Ion Torrent sequence reads were de-multiplexed , and each individual set of reads was aligned to reference sequences using Bowtie2 and SAMtools [78 , 79] . Aligned reads were visualized in the Integrative Genomics Viewer ( IGV ) [80] for single nt polymorphisms ( SNPs ) identification . Consensus sequence for each full-length genome was calculated using IGV . Read coverage ( reads/nt position ) was calculated using the genomecov command from BEDTools [81] . Sequence analyses were performed using MEGA v6 and Sequencher 5 . 4 ( Gene Codes Corporation , Michigan , USA ) . The consensus sequence was calculated using default settings in Sequencher v5 . 4 , and genomic sequences determined in this study were deposited into GenBank under Accession numbers KY424328 through KY424350 . All other relevant data are within the paper and its Supporting Information files . | Efforts are underway to develop vaccines against norovirus , a leading cause of acute gastroenteritis . The purpose of our study was to understand how norovirus strains within different genotypes evolve and adapt as they are transmitted in the human population . Using large-scale genomics and computational tools developed in our laboratory , we identified two strikingly different evolutionary patterns among norovirus genotypes: “static” and “evolving . ” We mined large datasets from infection and outbreak studies in context of these evolutionary patterns and propose a new model for antigenic clustering of norovirus genotypes that could simplify vaccine design . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"organismal",
"evolution",
"medicine",
"and",
"health",
"sciences",
"microbial",
"mutation",
"pathology",
"and",
"laboratory",
"medicine",
"genome",
"evolution",
"pathogens",
"variant",
"genotypes",
"microbiology",
"genetic",
"mapping",
"viruses",
"rna",
"viruses",
"phylogenetic",
"analysis",
"microbial",
"evolution",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"caliciviruses",
"genomics",
"medical",
"microbiology",
"norovirus",
"microbial",
"pathogens",
"molecular",
"evolution",
"evolutionary",
"immunology",
"molecular",
"biology",
"viral",
"evolution",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques",
"calicivirus",
"infection",
"heredity",
"viral",
"pathogens",
"virology",
"genetics",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"computational",
"biology",
"evolutionary",
"biology",
"organisms"
]
| 2017 | Static and Evolving Norovirus Genotypes: Implications for Epidemiology and Immunity |
Schistosomes cause more mortality and morbidity than any other human helminth , but control primarily relies on a single drug that kills adult worms . The newly transformed schistosomulum stage is susceptible to the immune response and is a target for vaccine development and rational drug design . To identify genes which are up-regulated during the maturation of Schistosoma mansoni schistosomula in vitro , we cultured newly transformed parasites for 3 h or 5 days with and without erythrocytes and compared their transcriptional profiles using cDNA microarrays . The most apparent changes were in the up-regulation of genes between 3 h and 5 day schistosomula involved in blood feeding , tegument and cytoskeletal development , cell adhesion , and stress responses . The most highly up-regulated genes included a tegument tetraspanin Sm-tsp-3 ( 1 , 600-fold up-regulation ) , a protein kinase , a novel serine protease and serine protease inhibitor , and intestinal proteases belonging to distinct mechanistic classes . The inclusion of erythrocytes in the culture medium resulted in a general but less pronounced increase in transcriptional activity , with the highest up-regulation of genes involved in iron metabolism , proteolysis , and transport of fatty acids and sugars . We have identified the genes that are up-regulated during the first 5 days of schistosomula development in vitro . Using a combination of gene silencing techniques and murine protection studies , some of these highly up-regulated transcripts can be targeted for future development of new vaccines and drugs .
The schistosome tegument , an unique double lipid bilayered syncitium that covers the external surface of the intra-mammalian developmental stages , represents the point of interaction between the parasite and mammalian host tissues . This structure is pivotal for parasite survival within the host and is therefore a primary target of anthelmintic drugs [1] and vaccines [2] , [3] . In similar fashion , the intestine , or gastrodermis of schistosomes is a source of secreted proteins and another point of interaction with host tissues ( i . e . blood ) . The Schistosoma mansoni genome sequence has recently been reported [4] and the secreted proteome ( secretome ) has also been characterised with a major focus on the proteins present in the tegument and excretory/secretory ( ES ) products [4] , [5] , [6] , [7] , [8] , [9] . While the schistosome gastrodermal proteome has not yet been explored , we recently described tissue-specific gene profiling for adult S . japonicum and characterised the transcriptome of gastrodermal cells using a combination of laser microdissection microscopy followed by cDNA microarray analysis [10] . Despite the progress made in characterising the mRNA and protein compositions of cells at the host-parasite interface , it is only now with the recent application of gene silencing technologies to the study of schistosomes , that we are understanding the functions of these proteins and how they enable schistosomes to exist as parasites [11] . In terms of vaccine development , the newly transformed schistosomulum is widely viewed as the most susceptible stage to antibody-mediated damage [2] , [3] , [12] , [13] , [14] . After cercariae transform into schistosomula , parasites undergo changes in their surface protein composition [15] . The schistosomula surface is dynamic , with some proteins appearing and others disappearing [16] as the parasites mature during their migration to the lungs . Once the parasites reach the lungs they are refractory to antibody-mediated damage [17] and cloak themselves in host blood group antigens [18] and other proteins involved in immune responses [19] , [20] . Obtaining sufficient quantities of schistosomula directly from lung tissue for most research purposes is time consuming and involves working with mammalian hosts . As a result , many researchers mechanically transform cercariae and culture them in serum-containing medium [21] . This in vitro strategy also confers uniformity in parasite maturation , which is critical for short culturing periods and cannot be achieved in vivo due to the variation in the time required for individual parasites to penetrate host skin and enter the vasculature . Erythrocytes are not usually included in the culture medium yet the parasites are bathed in blood in vivo . Moreover , schistosomula acquire erythrocyte [22] and other cellular [18] proteins onto their teguments , and this is thought to aid in evasion of the immune response . Despite their vascular existence , the effect of erythrocytes on gene expression of in vitro cultured parasites has not been addressed until now . We sought to identify the transcriptional changes in genes encoding surface and secreted proteins during the first 5 days of in vitro culture of schistosomula in the presence or absence of erythrocytes . Some of these surface exposed proteins are proving to be efficacious vaccines [23] , [24] , yet the expression profiles of only some of these genes have been explored , and have involved using arrays covering only ∼3 , 000 genes [25] . Here we show that the major transcriptional changes which occur during this 5 day time period involve a wide range of biological functions but prominent processes include tegument maturation , cellular development and organisation , gut function/nutrient acquisition and stress responses . The data provide a framework by which to select targets for vaccine and drug design based on genes that are critical for the development of S . mansoni larvae during their first few days in the mammalian host .
The Puerto Rican strain of S . mansoni and Biomphalaria glabrata snails were provided by the National Institutes of Allergy and Infectious Diseases Schistosomiasis Resource Centre at the Biomedical Research Institute ( Rockville , Maryland , USA ) . To obtain cercariae , B . glabrata snails infected with miracidia were exposed to incandescent light for 1 h and then washed twice in RPMI 1640 , 1% antibiotic/antimycotic and 10 mM Hepes ( Invitrogen ) . Cercariae were passed through a 22-gauge emulsifying needle 25 times to mechanically shear the cercarial tails from the bodies [26] . The resulting schistosomula were isolated from free tails by centrifugation through a 60% percoll gradient and washed three times with wash medium before experimentation [27] . Schistosomula were cultured at 37°C in modified Basch's medium ( containing 10% Fetal Calf Serum ) under 5% CO2 atmosphere [21] , [28] , either in the presence or absence of erythrocytes , for 3 h or 5 days . Parasites were washed in RPMI and stored in Trizol at −80°C until the total RNA was isolated following manufacturer's instructions . RNA quality , integrity checks and concentration were assessed using Nanodrop ND-1000 spectrophotometer and Agilent 2100 Bioanalyzer [29] . Due to the limited availability of schistosomula material , a single biological sample was used . The design and construction of the schistosome microarray has been previously reported [30] . The microarray comprises of 19 , 222 target sequences printed twice from two independent probe designs , including 12 , 166 probes derived from S . mansoni contiguous sequences ( contigs ) and 7 , 056 probes derived from S . japonicum contigs . The putative genes designated from S . mansoni contiguous sequences were the primary but not exclusive focus of the analysis , and the genes derived from S . japonicum contigs were also fully considered . Further details of the microarray design and the normalised data from this study are presented in Tables S1 and S2 . The methods used in microarray hybridisation and feature extraction have been previously reported [30] and followed the manufacturer's instructions ( One-Color Microarray-Based Gene Expression Analysis Protocol; Version 5 . 5 , February 2007 Agilent ) . For each sample 300 ng of total RNA was used to synthesise fluorophore-labelled cRNA using Cyanine 3-CTP ( Agilent Technologies One Color Microarray Kit ) . Samples were purified using the Qiagen RNeasy kit . Cyanine-labeled cRNA samples were examined at A260 and A550 using a ND-1000 spectrophotometer to determine yield , concentration , amplification efficiency and abundance of cyanine fluorophore . CY3c ( 1 . 65 µg aliquot ) was incubated with the fragmentation mix ( Agilent Technologies ) for 30 min at 60°C . Samples were then combined with 2× Gene Expression Hybridisation Buffer HI-RPM , mixed and applied to a gasket slide pre-positioned in a hybridisation chamber ( Agilent Technologies ) , placed in a hybridisation oven and incubated for 17 h at 65°C . Each hybridisation was performed in duplicate as a technical replicate . After hybridisation , microarray slides were washed using the standard protocol ( Agilent Technologies ) and scanned on an Agilent microarray scanner at 550 nm . The “tag image format files” ( tiff ) produced by the scanner were loaded into the image analysis program Feature Extraction 9 . 5 . 3 . 1 ( Agilent Technologies ) to establish standardised data for statistical analysis . All microarray slides were checked for background evenness by viewing the tiff image on Feature Extraction . Feature extracted data were analysed using GENESPRING software , version 7 . 3 . 1 ( Agilent Technologies/Silicon Genetics ) . Duplicated microarray data ( technical replicates ) were normalised using the GeneSpring normalisation scenario for “Agilent FE one-color” which included “Data Transformation: Set measurements less than 5 . 0 to 5 . 0” , “Per Chip: Normalise to 50th percentile” and “Per Gene: Normalise to median” . Samples were also normalised to individual samples depending on the comparison to be made . To compare developmental effects of the transformation from cercariae to schistosomula , 3 h schistosomula with erythrocytes and 5 day schistosomula with erythrocytes were normalised to expression levels of cercariae . To determine the effects of culturing schistosomula with erythrocytes , 3 h schistosomula with erythrocytes were normalised to 3 h schistosomula without erythrocytes; similarly 5 day schistosomula with erythrocytes were normalised to 5 day schistosomula without erythrocytes . Data sets were further analysed using published procedures [30] , [31] consisting of methods related to one-colour experiments and utilised the signal intensity ( gProcessedSignal ) values determined using Agilent's Feature Extraction software including aspects of signal/noise ratio , spot morphology and homogeneity . ProcessedSignal represents signal after localised background subtraction and includes corrections for surface trends . Features were deemed Absent when the processed signal intensity was less than two-fold the value of the processed signal error value . Features were deemed Marginal when the measured intensity was at a saturated value or if there was a substantial amount of variation in the signal intensity within the pixels of a particular feature . Features that were not Absent or Marginal were deemed Present . Data points were included only if Present and contigs were retained if all data points were Present . Transcripts that were at least two-fold up-regulated in a particular experiment were searched against a database consisting of the full protein datasets from the S . japonicum [32] and S . mansoni [4] genome sequencing projects using BLASTX . Protein hits with an identity greater than 95% were used as the protein translation of the EST and analysed using local versions of SignalP 3 . 0 [33] for secretory signal sequences and TMpred ( http://www . ch . embnet . org/software/TMPRED_form . html ) for transmembrane domains . Transcripts that had no high scoring identities in the predicted protein database were searched against the NCBI non-redundant protein database using BLASTX . If the top scoring protein had a bit score greater than 50 the reading frame of the blast translation was used to translate the EST into protein sequence . The protein sequence was then analysed using SignalP and TMpred . Gene expression patterns of a subset of genes were validated using real time PCR . Genes were chosen to span a range of biological functions and life cycle expression patterns , with a focus on including genes that encode known secreted/membrane proteins of interest as vaccines and drug targets , such as tetraspanins and intestinal proteases . Complementary DNA was synthesised from total RNA using a QuantiTect whole transcriptome kit ( Qiagen ) . All cDNA samples were synthesised from aliquots of the same total RNA and used for the microarray hybridisations at a concentration of 50 ng/µl quantified using a Nanodrop ND-1000 spectrophotometer . Subsequently , 1 µl aliquots were combined with 10 µl of SYBR Green , 3 µl of water and 2 µl ( 5 pmol ) of the forward and reverse primers in a 0 . 1 ml tube . All reactions were performed on a Rotor-Gene 3000 real time thermal cycler ( Corbett ) and analysed using Rotor Gene 6 Software ( Corbett ) . In order to minimise indiscriminate binding of double-stranded DNA , which can produce readings in the “no template” controls , separate reverse transcription and PCR steps were included . Primer sets used are described in Table S3 . The housekeeping gene TC15682 ( DNA segregation ATPase ) [34] was used for primary normalisation for all experiments . This housekeeping gene was selected from the initial microarray data since it was consistently unchanged throughout all of the comparisons made . Each experiment was performed in duplicate and the confidence threshold ( CT ) of the second set was normalised to the first set before evaluation . The analysis of correlation between microarray and quantitative PCR was performed in Graphpad Prism Version 5 ( Graphpad Software Inc . ) and was based on a previously published analysis [35] . To correlate results from microarray and quantitative PCR platforms , we first determined whether the data were distributed normally . This involved the use of both the “D'Agostino & Pearson omnibus normality test” and “Shapiro-Wilk normality test” . Both tests indicated that the data were not normally distributed; thus a Spearman correlation ( Rho ) was employed . All methods used an alpha value of 0 . 05 . Supplementary Information ( RAW Data ) has been submitted at GEO- Gene Expression Omnibus , ( http://www . ncbi . nlm . nih . gov/geo/ ) with accession numbers for the platform GPL7160 , and series GSE18335 .
Three hour and 5 day schistosomula were cultured in the presence of erythrocytes and the subsequent microarray data were normalised to the gene expression of cercariae . Data were filtered for each replicated data point for each of the 38 , 444 probes ( 19 , 222 contigs ) on the microarray . Data points were filtered to preserve signals that were flagged during the feature extraction process as Present in all hybridisations; this resulted in the retention of 13 , 466 probes ( 7 , 764 contigs ) . A final cut-off was applied to the microarray data generating lists of genes that were ≥2-fold up- or down- regulated ( relative to cercariae ) for both 3 h and 5 day schistosomula . The number of genes ( contigs ) that were ≥2-fold up-regulated in the 3 h and/or 5 day schistosomula were 1 , 608 at 3 hours and 3 , 600 at 5 days , with 1 , 270 genes maintaining up-regulation at both time points ( Figure 1A ) . Fewer genes were down-regulated ≥2-fold , with 1 , 183 at 3 hrs and 1 , 343 at 5 days; 831 genes remained down-regulated at both time points . The differential fold-change occurring between 3 h and 5 day cultured schistosomula was determined by plotting the signal intensity of the two time points and applying a >2 or <0 . 5 cut-off to represent a 2-fold up- or down- regulation . A larger number of genes were up-regulated during the transition from 3 h to 5 days ( 2 , 327 genes ) compared with those that were down-regulated during this time period ( 439 genes ) ( Figure 1B ) . Examples of differentially expressed genes with novel annotations as well as the expression levels of genes encoding surface proteins of recently described vaccine antigens are presented in Table 1 . Genes of particular interest that were highly up-regulated during the transition from cercariae to 3 h and 5 day schistosomula included the tegumental tetraspanin Sm-TSP-3 which was up-regulated almost 1 , 600-fold between cercariae and 5 day schistosomula . Other genes encoding tegument proteins that were upregulated included the Sm22 . 6 tegument associated antigen ( 137-fold upregulated at day 5 ) and a protein with homologues from the tegument of other flukes [36] and with low sequence identity to protein kinases ( TC7982 , 68-fold increased expression at day 5 ) . Genes encoding two additional tegument proteins were rapidly up-regulated at 3h ( annexin 7-fold and cytosol aminopeptidase 14-fold ) and then further up-regulated by day 5 ( annexin 21-fold and aminopeptidase 45-fold ) . A gene with sequence similarity to fasciclin 1 was up-regulated 100-fold in day 5 schistosomula ( TC14173 ) ; the full-length protein ( Smp_141680 ) was obtained from S . mansoni GeneDB ( http://www . genedb . org/genedb/smansoni/ ) and contained two predicted transmembrane domains . Genes encoding intestinal proteases with known or suspected roles in digestion of the blood meal were highly upregulated by day 5 and included cathepsin B ( 65-fold ) , cathepsin L ( 37-fold ) , cathepsin D ( 13-fold ) and cathepsin C ( 11-fold ) . The third most highly up-regulated gene in 5 day schistosomula was a S01 family serine protease ( TC16843; 108–150-fold ) , although its anatomical location , and therefore potential function , has not been determined . Some of the genes that were up-regulated very quickly after transformation ( by 3 h ) , included cathepsin B ( 10-fold ) , cathepsin L ( 2-fold ) and cathepsin D ( 2-fold ) . At a more global level , the transformation between cercariae and 3 h/5 day schistosomula resulted in up-regulation of genes encoding a wide range of gene ontology ( GO ) categories including stress effectors ( HSP70 and mitochondrial dicarboxylate transporter ) , enzymes involved in digestion of the blood meal ( proteases ) and iron storage ( ferritin 2 ( somal ferritin ) ; upregulated 135-fold ) ( Figure 2 ) . Other noteworthy genes included transmembrane transporters of zinc that were elevated in both 3 h and 5 day schistosomula ( TC7518 ZNT4_RAT Zinc transporter 4 , 3-fold at 3 h and 9-fold at 5 day; TC18440 solute carrier family 30 zinc transporter member 6 , 2- fold at 3 h and 4-fold at 5 days ) . Schistosomula cultured for 3 h or 5 days were maintained in the presence or absence of erythrocytes and the subsequent microarray data were normalised to the gene expression of parasites cultured without erythrocytes . Data were filtered for each replicated data point for each of the 38 , 444 probes ( 19 , 222 contigs ) on the microarray . Data points were filtered to preserve signals that were flagged during the feature extraction process as Present in all hybridisations; this resulted in the retention of 13 , 140 probes ( 7 , 599 contigs ) ( Figure 3A ) . A final cut-off was applied to the microarray data generating lists of genes that were ≥2-fold up- or down- regulated ( relative to culture conditions without erythrocytes ) for both 3 h and 5 day cultured schistosomula . A total of 788 genes at 3 h and 2 , 479 at 5 days were ≥2-fold up-regulated; 399 genes were up-regulated at both time points ( Figure 3B ) . A small number of genes ( 388 ) were down-regulated in 3 h schistosomula . A small proportion of genes that were up-regulated at 5 days were down-regulated initially at 3 h ( 317 genes ) . Examples of novel genes and gene whose expression was up-regulated during the development of schistosomula from the cercarial stage are presented in Table 1 , and further examples of novel up-regulated genes due to the presence of erythrocyte either at 3 h or 5 day schistosomula are presented in Table 2 . The presence or absence of erythrocytes had little to no effect on the expression of genes encoding intestinal proteases , and of the genes encoding exposed tegument proteins , only TSP-4 and alkaline phosphatase III were up-regulated ( 2-fold ) in the presence of erythrocytes . However , the inclusion of erythrocytes induced more obvious expression changes in genes involved in development and , to a lesser extent , the stress response ( Table 1 ) . These included genes encoding proteins involved in iron storage ( ferritin-2 , 20-fold increase ) , divalent metal transporter , heme-binding protein , proteolytic enzymes ( serine protease ) , a fatty acid receptor , glycogen formation ( glycogenin ) and other components potentially involved in cellular differentiation/morphology . An overview of the differentially expressed genes in the presence/absence of erythrocytes is presented as GO categories in Figure 4 . To validate the microarray transcription data , mRNA expression profiles for 15 genes from different functional categories were assessed using quantitative real time PCR . Two independent experiments were carried out for this validation . The relative differential gene expression obtained by microarray analysis and by quantitative PCR was similar for the majority of data points for the 15 genes assessed ( Figure 5 ) . There was a significant correlation of 0 . 9208 between the two data sets ( Spearman's Rho , P<0 . 0001 , n = 34 ) . Genes that were upregulated ≥2-fold in any of the comparisons were assessed for the presence of signal peptides/anchors and transmembrane domains , suggestive of an extracellular location and therefore potentially involved in host-parasite interactions . Eight percent of genes that were up-regulated in 3 h old schistosomula ( compared with cercariae ) encoded for secreted proteins; all of these had 20-fold or less increases in expression except for contig 680_298 , which was upregulated 3 , 447-fold in the absence of erythrocytes and an additional 742-fold in the presence of erythrocytes ( Figure 6 ) . Other genes up-regulated in 3 h parasites that encoded secreted proteins of interest included TC6882_676 , a serine protease inhibitor that was up-regulated 2-fold and then an additional 144-fold in 5 day old parasites . More than 9% of genes that were up-regulated in 5 day parasites ( compared to 3 hr parasites ) encoded for secreted proteins; fold-changes in expression were generally higher than those seen in 3 h parasites , with the top ten ranging from 1 , 598-fold ( TC18051_669 tetraspanin Sm-TSP-4 ) to 65-fold ( TC13586_1222 cathepsin B ) ( Figure 6 ) . Four of these top 10 most highly upregulated secreted proteins shared no identity with any proteins of known function . The complete list of genes encoding for secreted proteins that underwent ≥2-fold increased expression in any of the experimental groups is provided in Table S4 .
In this study we have presented a comprehensive analysis of the transcriptional changes that are associated with two distinct and critical phases of the early maturation of the intra-mammalian stages of S . mansoni . We investigated the developmental changes that occur in vitro ( 1 ) in the first few hours after transformation as the cercarial glycocalyx is shed and the parasite adapts to an intra-mammalian environment , and ( 2 ) as schistosomula mature in vitro over a 5 day period . This developmental window corresponds to the in vivo phase of parasite migration from the skin ( 3 h ) into the vasculature en route to the lungs ( 5 days ) . In vitro culture exposes the parasite to host serum ( and erythrocytes ) but does not entirely mimic intra-mammalian development due to the absence of tissue barriers to penetrate , a vascular system to navigate and a complete immune system to avoid . We also examined the effects on gene transcription when erythrocytes were present in the culture medium . For both analyses , we placed emphasis on genes that encode proteins at the host-parasite interface , namely exposed tegument proteins and key proteins involved in nutrient acquisition . The development of newly transformed S . mansoni schistosomula over the first 5–7 days as they enter the vasculature and progress to the lungs represents what many believe to be a critical window of opportunity for vaccine-mediated protection [3] , [37] , [38] . At this stage the parasite presents a distinct suite of proteins on its tegument and has not yet become fully cloaked in host-derived molecules [39] . Moreover , juvenile schistosomula are more susceptible to antibody-dependent cellular cytotoxicity [40] than are older schistosomula and adult worms . We therefore reasoned that genes that are highly expressed at this stage , particularly those encoding secreted and membrane proteins , are worthy targets for the development of vaccines and new drugs . Recent microarray based studies of S . mansoni [41] and S . japonicum [30] have profiled the gene expression patterns of a wide range of lifecycle stages . However , one aspect of the parasite lifecycle that has not been examined in detail for either species is the transformation of the cercarial stage to the “skin” schistosomula , and the maturation of these parasites into “lung” schistosomula . Dillon and co-workers [25] compared cDNAs from 2 day and 7 day old cultured schistosomula with a control pool of mixed lifecycle stage cDNAs . They utilised a cDNA microarray consisting of 3 , 088 unique contigs and identified broad categories of differentially expressed genes in the schistosomulum stage , including energy metabolism , cytoskeletal organisation , protease activity and chromosome remodelling . Our study differs from that of Dillon's in a number of ways; ( 1 ) we focused on different time points of cercarial transformation and maturation , 3 hours and 5 days post-transformation , since these times better approximate the skin and lung schistosomula stages of S . mansoni; ( 2 ) we compared the gene expression profiles of each developmental stage to the stage preceding it , i . e . schistosomula to cercariae instead of using a pooled multi-stage cDNA preparation for determining baseline expression; and ( 3 ) we utilized an array consisting of 12 , 166 S . mansoni unique contigs . Our results indicate that the transcriptional changes that accompany the development of cercariae to 3 hour and then to 5 day schistosomula are mostly found in the up-regulation of genes associated with many different molecular functions , but particularly catalytic activity and binding . Considerably fewer genes were down-regulated in the first 3 hours after transformation , most likely representing genes that are important for the free-swimming cercariae and are no longer required by the intra-mammalian stages . Jolly and co-workers found that the most highly up-regulated genes in cercariae ( compared to other life cycle stages ) were mitochondrial in nature representing the increased energy requirements needed for motility during this stage [41] . This observation was paralleled in our study where a number of mitochondrial genes such as NADH subunit 2 and NADH subunit 5 ( see Table S2 ) were down regulated ( up to 25-fold ) at 3 h post-transformation compared with the cercarial stage . We observed that most probes to actin or actin related genes and fibrillin 2 were upregulated in the schistosomula relative to cercariae . The differential expression of structural genes such as actin may reflect the increased levels of tegument matrix generation and turnover as well as smooth muscle formation in newly transformed and growing parasites [42] , [43] . We were particularly interested in the expression of genes encoding exposed proteins on the tegument of the parasite . Tetraspanins are four transmembrane spanning proteins represented by at least 5 distinct members in the tegument membranes of the adult parasite [2] , [5] , [6] , [44] . Sm-TSP-1 and TSP-2 are efficacious vaccines [24]; tsp-1 mRNA was up-regulated almost 4-fold in 5 day schistosomula but tsp-2 expression levels did not change . However , one of the most highly up-regulated gene on the entire microarray ( TC18051 ) , undergoing 1 , 598-fold up-regulation between 3 h and 5 day cultured parasites , was another tegument tetraspanin [6] that we have termed Sm-tsp-3 [2] . This considerable up-regulation of tsp-3 was confirmed by quantitative PCR . Genes encoding other tegument proteins that were significantly up-regulated during this developmental period included Sm22 . 6 , an unknown protein , and annexin . Sm22 . 6 is an inhibitor of human thrombin [45] , and its up-regulation upon entry into the host vasculature likely represents an important survival strategy . Recombinant Sm22 . 6 confers partial protection as a vaccine in murine studies [46] but it is a major target of IgE in infected people [47] , so its utility as a vaccine is likely limited . The third most highly up-regulated gene in 5 day schistosomula was a S01 family serine protease ( 150-fold ) , sharing identity with cercarial elastases that digest connective tissue proteins [48] , [49] . Unlike the cercarial elastases where mRNA expression is highest in sporocysts and is switched off in cercariae and intra-mammalian stages [50] , this new serine protease likely plays a distinct role due to its up-regulation in maturing schistosomula . We also detected 100-fold up-regulation in 5 day schistosomula of a gene encoding for a homologue of fasciclin 1 , a family of GPI- anchored proteins that mediate cell adhesion through an interaction with alpha3/beta1 integrin [51] . As schistosomula mature their gastrodermis forms and they begin to ingest blood as a source of nutrition . The digestive process begins with haemolysis where erythrocytes are ingested and lysed by the action of a haemolysin ( s ) within the oesophagus and intestine [52] . Saposin-like proteins ( SAPLIPs ) are candidate pore-forming haemolysins in the schistosome gut [53] , and we identified a SAPLIP that was highly up-regulated in 5 day schistosomula ( TC_14899 , 25-fold up-regulated ) . Two other SAPLIP-encoding genes , TC10647 and TC10646 , were also up-regulated ( 5- and 3-fold respectively ) in 5 day schistosomula , suggesting a role for this protein in blood feeding , either via haemolysis or transport of lipids . An anion sugar transporter , distinct from the well characterised glucose transporter family from the S . mansoni tegument [54] , was up-regulated 25-fold in 5 day schistosomula , although its site of expression ( gut , tegument or elsewhere ) has not yet been determined . Other intestinal genes that were up-regulated between 3 h and 5 day schistosomula were the digestive proteases . The role of cysteine and aspartic proteases in digestion of haemoglobin and serum proteins for nutritional support is well recognised [55] , [56] , and the increase in transcription of these genes as schistosomula begin to feed on blood supports their roles in this process . Erythrocytes provide the parasite with an important nutritional source . However , other host factors also have an impact on parasite development; for example , the presence of insulin in the culture medium induced the increased expression in S . japonicum of many genes related to sexual reproduction and protein translation in general [57] . In this current study , serum ( which contains insulin ) was present throughout during culture , but other components , such as glycoproteins or glycolipids , derived directly from the erythrocytes may have impacted on gene expression . The up-regulation of a low-density lipoprotein receptor in the presence of erythrocytes may be influential in the development of the female reproductive system since this protein is also a putative vitellogenin receptor ( Table 1 ) . Lipids are a component of the female reproductive tract in schistosomes , as indicated previously by the localisation of a fatty acid-binding protein within lipid droplets of vitelline cells [58] . The most commonly encountered transcripts that were up-regulated in the presence of erythrocytes were of retroviral/retrotransposon origin . This is not surprising , as we have previously reported the up-regulation of retrotransposons in geographical isolates of schistosomes [59] , and their up-regulation here in the presence of erythrocytes may be in response to environmental changes/stress . Other genes that were up-regulated in 5 day schistosomula in the presence of erythrocytes included at least three distinct mucins that were >3-fold up-regulated , and an activin receptor , which have been described from secretions of schistosome eggs and cercariae [60] and the tegument [61] . Glycosidases were also upregulated , an example being glucan 1 4-beta-glucosidase ( TC16784 ) . This was 2 . 8 fold up-regulated in 5 day schistosomula in the presence of erythrocytes compared to their erythrocyte-free cultured counterparts , further implying a general increase in expression of genes involved in acquisition and metabolism of erythrocyte proteins , lipids and glycans . Only one of the digestive proteases showed greater than 2-fold increased expression in the presence of erythrocytes ( TC10495 , homologue to cathepsin B1 ) , implying that their up-regulated expression is predominantly independent of erythrocytes ( and therefore haemoglobin ) , but instead is dependent on serum proteins . Selected examples of the most highly up-regulated ( non-retroviral ) genes in the presence of erythrocytes in 5 day schistosomula are provided in Table 2 . The transformation from cercaria to schistosomulum requires the adaption of the parasite to radically different environments and subsequent large scale cellular differentiation and growth . This stressful process is reflected by the up-regulation of multiple stress related genes [62] , including numerous heat shock proteins and minichromosomal maintenance complex component 2 , all of which were up-regulated at 3 h and were maintained at elevated expression levels through day 5 , relative to cercariae . Another major feature of the maturing schistosomula is the extensive musculature that begins to form beneath the tegument [42] . This increase in muscle tissue was reflected in our study by increased expression levels of a number of muscle related genes including paramyosin , myosin light chain kinase and myosin light polypeptide 5 regulatory sequence ( See Table S2 ) . Paramyosin is of particular importance to intra-mammalian stages due to its dual function as a structural element of smooth muscle and its immunomodulatory function in schistosomula [63] , [64] , [65] . The tegument of S . mansoni schistosomula is the major target of the immune response of a population of putatively resistant ( PR ) individuals in S . mansoni endemic areas of Brazil [66] , and in schistosomiasis resistant rats [67] . By identifying genes that are highly upregulated during this developmental process , particularly those that are accessible to antibodies such as the exposed tegument and gut proteins , a new suite of vaccine antigens can be produced and screened with sera from resistant hosts . We envisage this study as an initial screen for potential vaccine antigens and drug targets , to be followed by murine protection studies and functional analysis where approaches such as RNA interference could be used to verify the consequences of silencing these genes under different in vitro and in vivo conditions . | Schistosome blood flukes cause more mortality and morbidity than any other human worm infection , but current control methods primarily rely on a single drug . There is a desperate need for new approaches to control this parasite , including vaccines . People become infected when the free-swimming larva , the cercaria , enters through the skin and becomes the schistosomulum . Schistosomula are susceptible to immune responses during their first few days in the host before they become adult parasites . We characterised the genes that these newly transformed parasites switch on when they enter the host to identify molecules that are critical for survival in the human host . Some of these highly up-regulated genes can be targeted for future development of new vaccines and drugs . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
]
| [
"genetics",
"and",
"genomics/gene",
"expression"
]
| 2010 | Transcriptional Changes in Schistosoma mansoni during Early Schistosomula Development and in the Presence of Erythrocytes |
Acinetobacter baumannii , A . nosocomialis , and A . pittii have recently emerged as opportunistic human pathogens capable of causing severe human disease; however , the molecular mechanisms employed by Acinetobacter to cause disease remain poorly understood . Many pathogenic members of the genus Acinetobacter contain genes predicted to encode proteins required for the biogenesis of a type II secretion system ( T2SS ) , which have been shown to mediate virulence in many Gram-negative organisms . Here we demonstrate that Acinetobacter nosocomialis strain M2 produces a functional T2SS , which is required for full virulence in both the Galleria mellonella and murine pulmonary infection models . Importantly , this is the first bona fide secretion system shown to be required for virulence in Acinetobacter . Using bioinformatics , proteomics , and mutational analyses , we show that Acinetobacter employs its T2SS to export multiple substrates , including the lipases LipA and LipH as well as the protease CpaA . Furthermore , the Acinetobacter T2SS , which is found scattered amongst five distinct loci , does not contain a dedicated pseudopilin peptidase , but instead relies on the type IV prepilin peptidase , reinforcing the common ancestry of these two systems . Lastly , two of the three secreted proteins characterized in this study require specific chaperones for secretion . These chaperones contain an N-terminal transmembrane domain , are encoded adjacently to their cognate effector , and their disruption abolishes type II secretion of their cognate effector . Bioinformatic analysis identified putative chaperones located adjacent to multiple previously known type II effectors from several Gram-negative bacteria , which suggests that T2SS chaperones constitute a separate class of membrane-associated chaperones mediating type II secretion .
Members of the genus Acinetobacter are regarded as opportunistic human pathogens of increasing relevance worldwide due in part to the rapid emergence of multiply-drug resistant strains [1] . In fact , the Center for Disease Control and Prevention has recently categorized multi-drug resistant Acinetobacter at the serious hazard level , prompting sustained research and action to further prevent its dissemination . Specifically , A . baumannii , A . pittii , and A . nosocomialis of the Acinetobacter calcoaceticus-baumannii ( Acb ) complex have become the most medically relevant members of the genus as they are most frequently isolated from health care facilities as well as human tissues [2] . Although A . baumannii is thought to be the most prevalent and virulent member of the genus Acinetobacter , both A . pittii and A . nosocomialis are capable of causing severe human disease and are likely under-represented due largely to technological limitations in species identification across clinical laboratories worldwide [3–5] . The ability of Acinetobacter to persist in health care facilities has been an active area of investigation; however , it has been mostly limited to the mechanisms utilized to resist antimicrobial therapy , desiccation , and disinfectants . Little is currently known about the virulence factors employed by Acinetobacter species ( spp . ) to colonize and infect different human tissues [6–9] . Recent studies have , however , demonstrated that protein glycosylation [10 , 11] , capsule production/modulation [12–14] , metal acquisition strategies [15 , 16] , outer membrane proteins [17–19] , and alterations in lipid A [8] , all contribute to the ability of medically relevant Acinetobacter species to cause disease . It has also been shown that Acinetobacter spp . produce both type I pili and type IV pili; however , a definitive role for these pili in virulence has not been determined [20–22] . Multiple secretion systems have been identified and characterized for their role in the biology and virulence of medically relevant members of the Acb . The most comprehensively studied secretion system in Acinetobacter is the type VI secretion system ( T6SS ) , which has been functionally identified and studied in the medically relevant species A . nosocomialis and A . baumannii , as well as in the non-pathogenic species A . baylyi [23 , 24] . Recently , it was found that several multidrug resistant strains of A . baumannii carry a large , self-transmissible plasmid that encodes for the negative regulators of T6SS . It was found that T6SS is silenced in plasmid-containing cells while part of the population loses the plasmid and subsequently activates T6SS [25] . However , unlike Burkholderia pseudomallei , which utilizes its T6SS to toxically infect eukaryotic cells [26 , 27] , the Acinetobacter T6SS primarily mediates anti-bacterial killing; yet , a recent study identified the Acinetobacter T6SS to be required for full virulence in an insect model [28] . A type V system autotransporter , Ata , has also been characterized and found to mediate biofilm formation , adherence to extracellular matrix proteins , as well as virulence in a murine systemic model of Acinetobacter infection [29] . Furthermore , plasmid encoded genes required for the biogenesis of a type IV secretion system ( T4SS ) in A . baumannii and A . lwoffii have been bioinformatically identified [30 , 31]; however , no empirical evidence demonstrating their function has been presented . To date , no classical toxins have been described nor have any bona fide secretion systems specifically related to disease been discovered in medically relevant Acinetobacter members . Genes encoding proteins predicted to be associated with a type II secretion system ( T2SS ) have been identified in A . baumannii [32 , 33] . T2SS are multi-protein complexes , evolutionarily related to type IV pili ( T4P ) systems , which are responsible for the export of proteins from the periplasmic space to the extracellular milieu or to the outer surface of many Gram-negative bacteria [34 , 35] . The T2SS is composed of 12–15 proteins comprising four sub-assemblies: a pseudopilus , an inner-membrane platform assembly , an outer-membrane complex , and a secretion ATPase [36] . Effector proteins are first translocated to the periplasm by the general secretory ( Sec ) pathway or the twin arginine transport ( Tat ) system , where the targeted proteins can then fold into the correct tertiary and/or quaternary structure prior to association with components of the T2SS [37] . Competently folded effector proteins can then interact with the different subassemblies of the T2SS and be extruded via interactions with the pseudopilus and the outer-membrane secretin [38] . Several Gram-negative pathogens , including Vibrio cholerae [39 , 40] , Legionella pneumophila [41 , 42] , and enterotoxigenic Escherichia coli [43] , utilize T2SS for the export of toxins as well as proteins associated with the degradation of biopolymers; thus , T2SS can serve both pathogenic and survival roles for bacteria depending on the environmental niche . Here , utilizing a proteomics approach coupled with mutational analyses , we demonstrate that Acinetobacter spp . carry a functional T2SS . We also present the type II secretome of A . nosocomialis strain M2 . Using a mutational analysis approach , we further demonstrated that both the type IV pili system and the T2SS share a common prepilin peptidase , PilD . Importantly , we show that two of the three identified effectors required chaperones for secretion by the T2SS , one of which is a newly characterized protease/chaperone pair . Lastly , we demonstrated that the Acinetobacter T2SS contributes to the extracellular lipolytic activity , and the virulence in the both the Galleria mellonella infection model and murine pulmonary infection model .
Previous manuscripts have reported the bioinformatic identification of genes predicted to encode proteins required for the biogenesis of a T2SS in Acinetobacter spp . [32 , 33] . We have also identified homologs of genes associated with the biogenesis of a T2SS in A . nosocomialis strain M2 . Here we adopt the gsp nomenclature for general secretory pathway when defining homologous T2SS associated genes in Acinetobacter . Using the Basic Local Alignment Search Tool ( BLAST ) [44] and homologs of known T2SS-associated genes from V . cholerae , P . aeruginosa , and E . coli , we identified several gsp homologs in all publically available genomes from medically relevant Acinetobacter spp . Unlike many Gram-negative pathogens encoding a T2SS , the genes encoding predicted type II secretion biogenesis proteins were not encoded in a single operon [35] , but were grouped into five distinct gene clusters separated over large distances on the chromosome ( Fig 1 ) . To test the functionality of the T2SS in A . nosocomialis strain M2 we deleted the predicted type II outer membrane secretin gene homolog , gspD , from strain M2 . GspD secretin monomers form a dodecamer complex in the outer-membrane that is required for the export of periplasmic effector proteins ( Fig 1 ) [46 , 47] . Using the T2SS deficient M2∆gspD::kan mutant we probed for differentially secreted proteins by one-dimensional sodium dodecyl sulfate polyacrylamide gel electrophoresis ( SDS-PAGE ) . Furthermore , we complemented the gspD::kan mutant and probed for secreted proteins from this genetic background . The secreted protein profiles from all three strains contained an abundance of proteins; however , differences in the secreted protein profile from the gspD::kan mutant were clearly evident when compared to the parental strain . At least 4 silver-reactive protein bands were absent in the secreted profile from the gspD::kan mutant when compared to the secreted protein profile from the parental strain ( Fig 2A ) . Importantly , the secreted protein profile from the complemented gspD strain showed the same profile as the parental strain M2 indicating that these differences observed in the secreted protein profile from the gspD::kan mutant were due to the loss of the putative outer membrane secretin and not to the mutational strategy . Although our 1D SDS-PAGE analysis strongly indicated that A . nosocomialis strain M2 did in fact produce a functional T2SS , the abundance of non-type II secreted proteins would interfere with downstream identification . We therefore proceeded with a two-dimensional difference gel electrophoresis ( 2D-DIGE ) analysis to enhance protein separation . The secreted protein fraction from the wild type strain M2 was compared to the secreted protein fraction from the M2∆gspD::kan mutant to generate the preliminary type II secretome of A . nosocomialis strain M2 via 2D-DIGE analysis . Analysis of gel images with SameSpots software ( TotalLab , New Castle upon Tyne ) revealed that 60 spots exhibited a statistically significant average change of at least 4-fold when comparing wild type M2 vs . M2∆gspD::kan samples . A representative gel image from the 2D-DIGE analysis is shown in Fig 2B . Gel spots were cored using an Ettan Spot Handling Workstation and prepared for in gel trypsin digestion . Peptides were eluted and analyzed using capillary-liquid chromatography-nanospray tandem mass spectrometry . The complete list of proteins identified for each spot as well as a detailed description of the 2D-DIGE analysis and methodologies can be found in S1 Appendix; proteins associated with the largest spot fold change , however , are listed in Fig 2C . Three of the proteins identified in Fig 2C , M215_05100 , M215_10380 , and M215_03235 , were of particular interest as all contained domains of known function . The remaining proteins listed in Fig 2C do not contain any known functional domains , with the exception of M215_02250/M215_02255 pair , which was bioinformatically identified as GlyGly-CTERM and rhomobosortase [48] . The top secreted candidate , M215_05100 , is an ortholog of the previously identified CpaA metallopeptidase from the M72 family of peptidases , which was proposed to cleave both factor V and fibrinogen [49] . The M72 peptidases are characterized as peptidyl-Asp-endopeptidases containing the HEXXHXXGXX active site , where a zinc ion is predicted to be bound by three histidine residues , and the glutamate is predicted to be the catalytic residue [50] . The M215_10380 locus encodes an ortholog of the previously characterized LipA lipase from A . baylyi [51 , 52] , which contains an alpha/beta hydrolase fold from the homologous family abH15 . 02 ( B . cepacia lipase-like ) within the abH15 superfamily ( Burkholderia lipase superfamily ) as determined by the Lipase Engineering Database [53] . These lipases are predicted to have a catalytic triad of a serine , a glutamate or aspartate , and a histidine . Lastly , the M215_03235 locus encodes for another protein containing an alpha/beta hydrolase fold; however , the M215_03235 gene product does not have homology to any know lipases within the Lipase Engineering Database and has yet to be characterized in Acinetobacter . BLAST analysis revealed the presence of only a single prepilin peptidase , gspO/pilD , which was previously designated PilD and reported to be the major prepilin peptidase for the type IV pili ( T4P ) system in Acinetobacter ( Fig 1 ) [45] . Given that only one gspO/pilD homolog was identified in strain M2’s genome as well as in A . baumannii ATCC 17978 and 19606 , we hypothesized that the previously identified prepilin peptidase , PilD , was also the pre-pseudopilin peptidase required for the T2SS . To this end , we cloned and heterologously expressed the predicted major pseudopilin , gspG , with a carboxy-terminal FLAG tag in the wild type M2 background , the ∆pilD::kan mutant , and its respective complement in order to probe for pseudopilin processing . As expected GspG-FLAG expression was detected in all three backgrounds; however , GspG-FLAG from both the wild type M2 and the complemented pilD strain migrated with an increased electrophoretic mobility as compared to GspG-FLAG from the ∆pilD::kan strain ( Fig 3 ) . The increase in electrophoretic mobility was most likely due to the loss of the leader sequence of GspG; furthermore , PilD was required for the processing observed . Lastly , an additional band of intermediate electrophoretic mobility was detected only in the ∆pilD::kan background . We hypothesize this form of GspG-FLAG to be a degradation product . Our 2D DIGE analysis indicated that the CpaA metallopeptidase was secreted via the T2SS; therefore , we used an immunoblotting approach to verify CpaA secretion was type II dependent . We cloned the cpaA gene with its predicted native promoter into the Acinetobacter-E . coli shuttle vector pWH1266 [54] , containing a hexa-histidine tag onto the carboxy terminus of cpaA . Hexa-histidine tagged CpaA was expressed in trans in multiple genetic backgrounds to probe for expression and secretion . CpaA-His expression was detected in all strains tested , however , it was only detected in the secreted fractions from strains predicted to have a fully functioning T2SS ( Fig 4C ) . Specifically , neither the ∆gspD::kan mutant nor the ∆pilD::kan mutant secreted CpaA-His , indicating the dependency of the T2SS for active export of CpaA . As expected secretion was independent of the type IV pilus as the ∆pilA::frt mutant displayed active secretion of CpaA-His . Immediately downstream of cpaA is the M215_05105 open reading frame , which when analyzed by BLASTp did not identify any known functional domains . However , when the M215_05105 ORF was analyzed by Domain Enhanced Lookup Time Accelerated ( DELTA ) BLASTp , which has higher sensitivity than BLASTp [55] , the M215_05105 ORF was found to contain a domain from the SRPBCC superfamily ( Fig 4A and 4B ) . Proteins carrying a domain from the SRPBCC superfamily are predicted to contain a deep hydrophobic ligand-binding pocket and have chaperone-like activity [56] . We thus hypothesized that the M215_05105 gene product , designated CpaB due to its proximity to CpaA , was a CpaA-specific chaperone . To test our hypothesis , we deleted the cpaB gene and probed for CpaA-His expression and secretion . As shown in Fig 4D , CpaA-His expression was detected in the ∆cpaB::frt mutant; however , CpaA-His was not secreted , indicating that CpaB was required for CpaA secretion . Importantly , we were able to reintroduce the cpaB allele and restore the active secretion of CpaA-His . To further demonstrate the dependency of CpaA secretion on CpaB , we heterologously expressed cpaA-his alone or in tandem with cpaB in A . baumannii ATCC 19606 , which does not encode for orthologs of either the CpaA metallopeptidase or the CpaB chaperone , yet is predicted to produce a functional T2SS . As shown in Fig 4E , CpaA-His was expressed but not secreted by 19606 cells when the pWH-cpaA-his plasmid was introduced , however , when both cpaA-his and cpaB were co-expressed , CpaA-His was secreted , indicating that CpaA secretion is not only dependent on a functional T2SS , but also on the chaperone activity of CpaB . The M215_10380 ORF , encoding for a LipA ortholog , was also identified in our 2D-DIGE analysis as a type II effector . It has been previously demonstrated in A . baylyi and Pseudomonas that secretion and over-expression of LipA orthologs are dependent on a LipB-like chaperone [51 , 57] . In A . nosocomialis M2 , a lipB homolog is adjacent to lipA ( Fig 5A and 5B ) . When LipA-His was over-expressed from the pWH-lipA-his plasmid , we did not detect its secretion . However , when we co-expressed the upstream lipB gene with lipA-his , LipA-His was expressed and secreted in all backgrounds predicted to have a functional T2SS ( Fig 5C ) . LipA-His was neither detected in the secreted fraction from the ∆gspD::kan mutant nor the ∆pilD::kan mutant . Secretion was also independent of the type IV pilus fiber itself ( Fig 5D ) . We also confirmed that LipA was secreted in a LipB chaperone-dependent manner by A . baumannii ATCC 19606 ( Fig 5E ) . To confirm that LipA is in fact a lipase , we purified culture supernatants from multiple genetic backgrounds and probed for lipolytic activity as determined by a modified para-nitrophenol palmitate ( p-NPP ) assay [58] . As seen in Fig 5F , culture supernatants from the wild type M2 exhibited lipolytic activity as demonstrated by an increase in the absorbance at 410nm ( A410 ) over a 12-hour time period . Culture supernatants from the ∆gspD::kan mutant displayed only minimal increases in the A410 indicating almost a complete lack of lipase activity . Importantly , the complemented gspD strain displayed very similar increases in the A410 when compared to the wild type , indicating that the lipase activity in culture supernatants was mainly dependent on the T2SS . Culture supernatants from the lipA::kan mutant exhibited an approximately 50% reduction in lipase activity; furthermore , the complemented lipA strain regained activity; in fact , culture supernatants from the complemented lipA strain displayed approximately a 30% increase in lipase activity over the wild type strain . Next we purified culture supernatants from the lipB::frt mutant and found that it displayed the same profile as the lipA mutant when measuring the A410; however , when we reintroduced the lipB gene into the lipB::frt mutant , we observed minimal complementation ( Fig 5F ) . The 2D-DIGE analysis revealed that the spot corresponding with the M215_03235 protein was associated with an 8 . 1 fold change when compared to the ∆gspD::kan mutant . The M215_03235 gene encodes for a protein containing multiple predicted domains including a LIP domain ( pfam03583 ) , a DAP2 domain ( COG1506 ) , and two AB hydrolase_5 domains ( pfam 12695 ) . Given that all of these domains are associated with predicted lipase/esterase activity , we have designated M215_03235 as lipH in order to avoid confusion with previously characterized lipases . To confirm that LipH was secreted in a T2SS-dependent manner , we utilized a similar approach as described above where we cloned and tagged LipH into pWH1266 with a carboxy-terminal his tag . We then introduced this construct into multiple strains and probed from LipH-His expression and secretion . As seen in Fig 6A , LipH-His was detected in whole cell lysates of all strains tested; however , LipH-His was found to only be secreted in strains predicted to express a functional T2SS . We further assessed the ability of a panel of clinical isolates to secrete LipH-His . As shown in Fig 6B , LipH-His expression and secretion was detected in all clinical isolates tested . Because alpha/beta hydrolase domains , such as the one present in LipH , are commonly found in lipases , we verified that LipH has lipolytic activity . We constructed a ∆lipH::kan mutant as well as a lipH complemented strain and subjected these strains to the p-NPP assay utilized above for LipA . As seen in Fig 6C , the ∆lipH::kan mutant displayed an increase in the A410 , indicating lipolytic activity; however , the increase was substantially lower than both the parent strain as well as the lipH complemented strain indicating that the LipH protein is a lipase . The greater wax moth , Galleria mellonella , has been routinely used to assess the virulence of Acinetobacter [59] . Furthermore , strains with attenuated virulence in the G . mellonella model have also been shown to have attenuated virulence in murine models of infection [60] . In order to assess the role of the Acinetobacter T2SS in the G . mellonella model , we first determined the LD50 for the wild type A . nosocomialis strain M2 . Groups of ten larvae were each injected with 10μL of either approximately 105 , 106 , or 107 total CFU of strain M2 , incubated at 37°C for 24 hours , and checked for viability as determined by accumulation of melanin and loss of movement . From these studies , the LD50 was determined to be approximately 3X106 CFU and was selected as the inoculation dose for subsequent infections ( S1 Fig ) . The wild type M2 , gspD::kan mutant , and the gspD complemented strain were individually injected into cohorts of G . mellonella at the specified dose , incubated at 37°C for 24 hours and checked for viability . As expected , 50% of the larvae injected with either the wild type M2 or complemented gspD strain succumbed to the infection ( Fig 7A ) ; however , only 30% of the larvae injected with the M2∆gspD::kan mutant died after 24 hours . To further demonstrate that the Acinetobacter T2SS contributes to the virulence of Acinetobacter in the G . mellonella model , we injected cohorts of larvae with the pre-determined LD50 for the M2∆gspD::kan mutant of 107 CFU ( S2 Fig ) . As seen in Fig 7B , 50% of the larvae injected with M2∆gspD::kan mutant died at the specific dose; however , 80% of the larvae injected with the wild type M2 died as a result of the infection after 24 hours . Interestingly , almost all of the larvae ( ~97% ) injected with the complemented gspD strain died after 24 hours . Acinetobacter infections most frequently manifest as pneumonias , specifically , within the mechanically ventilated patient population [61] . The murine acute pulmonary infection model has therefore been developed to model an active Acinetobacter pneumonia clinical presentation . In order to determine a role of the T2SS in Acinetobacter virulence , we first constructed a strain of A . nosocomialis with an unmarked , in-frame deletion of gspD , which encodes for the predicted outer-membrane secretin . Prior to infection studies , we verified that the newly generated M2∆gspD::frt mutant was in fact impaired in secretion of type II effector proteins ( S3 Fig ) . Using our previously described murine infection model [62] , we performed infection experiments with either the wild type A . nosocomialis strain M2 , the unmarked , isogenic M2∆gspD::frt mutant , or its respective gspD complemented strain . Mice were intranasally inoculated with 1X109 CFU , as we previously determined that inoculating mice with this dose of wild type bacteria resulted in full murine viability , yet , resulted in significant organ-specific bacterial burden ( S4 Fig ) . Groups of mice were individually administered an intransal inoculation of either the wild type strain , the ∆gspD::frt mutant , or the respective complemented gspD strain . Thirty-six hours post-infection , mice were sacrificed and the lungs , spleen , and livers were harvested in order to determine total bacterial burdens . As seen in Fig 8A , mice infected with either the wild type strain or the complemented gspD strain all had high bacterial burdens in the lungs . Furthermore , bacterial burdens displayed limited variability indicating a full level of complementation for the gspD complementation strain . Mice infected with the ∆gspD::frt mutant displayed significantly lower bacterial burdens in the lung when compared to either the wild type or complemented gspD strain . A similar trend was also observed for bacterial burdens in the spleen , where , mice infected with either the wild type or the complemented gspD strain had significantly higher bacterial burdens ( Fig 8B ) . We also enumerated bacterial colony forming units from the livers of infected mice and did not observe any significant differences between the cohorts ( Fig 8C ) . We have shown that two of the three secreted type II effectors identified in A . nosocomialis strain M2 require specific chaperones for secretion . To date only the lipase-specific foldases ( Lifs ) have been characterized as chaperones for type II effectors [51 , 63–65] . Indeed , a complex of the B . glumae LipA/Lif has been crystallized [66] . The Lifs are unique steric chaperones , which have an N-terminal membrane-anchor and a C-terminal domain that facilitates proper folding of their cognate lipase upon entry into the periplasm [67] . Furthermore , the first characterization of a chaperone participating in the secretion of a type II secreted protein from Acinetobacter was described in 1995 . These authors demonstrated that a lipase specific chaperone , designated LipB , was required for secretion of the LipA lipase . They found that the C-terminal domain of the LipB chaperone was located outside of the cytoplasm . Lastly , in contrast to what had been previously found in Pseudomonas strains , the authors found that lipB was actually encoded upstream of lipA [51 , 52] . We have expanded upon this paradigm with the identification of a novel protease/chaperone pair ( CpaA/B ) . Furthermore , we hypothesized this phenomenon to be more widespread . In order to identify putative chaperones of type II secreted proteins , we first searched for open reading frames ( ORFs ) encoded adjacently to known type II effectors that were predicted to be part of the same operon . We then narrowed our search to ORFs that encode for proteins with a predicted N-terminal transmembrane domain as this feature is shared both by the Lifs and the newly characterized CpaB chaperone . As found in Table 1 , we were able to identify several putative chaperones of type II effectors in diverse Gram-negative bacteria such as V . cholerae , P . aeruginosa , and B . pseudomallei , which suggests that CpaB , LipB , and Lifs belong to a family of membrane-bound chaperones involved in T2SS secretion .
Acinetobacter spp . have rapidly emerged as significant opportunistic pathogens afflicting healthcare facilities worldwide . Although sophisticated studies track the epidemiology of outbreaks worldwide , our collective understanding of the molecular mechanisms employed by Acinetobacter spp . to cause disease is in its infancy . In this work , we combined bioinformatics , proteomics , mutational analyses , and virulence assays to demonstrate that Acinetobacter spp . produce a functional T2SS , which is required for the secretion of multiple proteins that are required for full virulence . Importantly , this is the first bona fide secretion system required for virulence in a mammalian model identified in Acinetobacter . Notably , two of the three secreted proteins characterized in this study require dedicated chaperones for type II secretion . While this paper was under revision , an article reporting the presence of a functioning T2SS in A . baumannii ATCC 17978 was published [76] . In this work , it was found that 17978 also required a T2SS for the secretion of the LipA lipase and growth on minimal media with olive oil as the sole carbon source . It was also found that both the 17978∆gspD and 17978∆lipA mutants were less fit in a murine septicemia model when competed against the parental strain . Typically , T2SSs secrete as many as 18–25 proteins and facilitate the delivery of major virulence factors to the extracellular environment for many important human pathogens , such as Legionella pneumophila and V . cholerae [40 , 42] . Here , we utilized the 2D-DIGE method coupled with mutational analyses to characterize the type II secretome for A . nosocomialis strain M2 . Our analysis identified over 60 spots with a 4-fold difference when comparing the wild type M2 vs . M2∆gspD::kan mutant; however , we concentrated our efforts on three proteins that contain domains of known functions . Other studies will be needed to determine the role of the remaining type 2 effector candidates and of individual secreted proteins in Acinetobacter pathobiology given the importance of this system in virulence . The genetic architecture of T2SSs usually consists of between 12 and 15 genes , most of which appear to be organized in a single operon [35] . From a regulatory standpoint , the single operon arrangement of T2SS associated genes would seem to be the simplest to transcriptional control . However , as noted above , the T2SS associated genes from A . nosocomialis strain M2 are found in five distinct genetic loci , a genetic arrangement that resembles the type IVa pilus system [77] . Furthermore , this dispersed genetic arrangement is highly conserved across different Acinetobacter species , including the pathogenic species A . baumannii and the non-pathogenic species A . baylyi . Closer examination of each T2SS gene cluster does not provide any obvious insights into the regulatory mechanisms as some T2SS genes appear to be in putative operons with other genes not known to be associated with T2SSs . Outside of the genus Acinetobacter the same genetic architecture can also be found in bacteria from the genus Psychrobacter . As demonstrated previously , the prepilin peptidase PilD was required for major pilin processing and proper functionality of T4P in A . nosocomialis strain M2 [45] . Our current data demonstrated that PilD is also required processing of the predicted major pseudopilin , GspG , and thus secretion of T2S substrates . Given the strong evolutionary relatedness between the T4P system and the T2SS , the phenomenon of sharing protein components between two functionally distinct systems does not seem impractical , nevertheless , it is uncommon . To date only D . nodosus [78] , P . aeruginosa [79] , V . cholerae [80] , and L . pneumophila [81 , 82] have been demonstrated to share a prepilin peptidase between both the T4P system and a T2SS . Of the three type II effectors studied , only LipA has previously characterized orthologs , which were primarily described in Pseudomonas and also require a chaperone [83] . However , to date , none of these lipases have been connected to pathogenesis . We demonstrated that the LipA lipase was responsible for approximately half of the lipase activity observed from the secreted fraction of the wild type strain M2 . As expected , LipA activity was also dependent on the LipB chaperone , as supernatants from the ∆lipB::frt mutant displayed nearly identical lipase activity levels as the ∆lipA::kan mutation . However , our lipB complemented strain only marginally increased the lipase activity of the ∆lipB::frt mutant , indicating that even though we constructed an in-frame , unmarked mutation in the lipB gene , we may still be observing polar effects on lipA transcription . The lipA gene is 81bp downstream of the lipB gene and therefore could potentially have its own promoter that is partially contained within the 3’ region of the lipB gene . We and others have observed similar cryptic promoter events during previous studies of the pilTU gene cluster , where an in-frame , unmarked mutation of pilT still had polar effects on pilU expression [45 , 84] . Even in the absence of lipA , culture supernatants retained residual lipase activity as compared to the gspD mutant strain . As such , we found that LipH mediated lipase activity of culture supernatants as well . A BLASTp search of LipH orthologs outside of Acinetobacter identified similar proteins found in bacteria from the genus Myriodes , some of which act as opportunistic human pathogens [85] , as well as bacteria from the genus Bacillus; however , none of those orthologs have been characterized . Using A . nosocomialis strain M2 as our model system we demonstrated that LipH secretion was indeed dependent on a functional T2SS . We also demonstrated that T2SS is conserved and functional across Acinetobacter spp . via immunoblotting of epitope tagged effectors . Specifically , we showed that LipH from M2 was secreted by a panel of Acinetobacter clinical isolates , including , A . calcoaceticus , A . baumannii , A . pittii , and A . junnii . We also demonstrated that A . baumannii ATCC 19606 could secrete both LipA and CpaA; however , as expected the respective chaperones for each protein were required for active secretion . These data strongly suggest the presence of a functional T2SS in the majority of medically relevant Acinetobacter spp . This hypothesis is further supported by the fact that genes predicted to encode proteins required for the biogenesis of the T2SS are highly conserved and distributed amongst Acinetobacter spp . The remaining effector characterized in our study was the CpaA metallopeptidase . CpaA was previously purified from culture supernatants [49]; however , its mechanism of secretion was not determined . It was previously shown that CpaA is involved in degradation of Factor V and fibrinogen , which would result in a decrease in clotting activity . Here , we demonstrated that CpaA was secreted in abundance in a type II dependent manner , yet , was also dependent on a novel chaperone , designated CpaB . CpaB is the first characterized T2SS chaperone devoted to the secretion of a protease . Topological modeling of the CpaB chaperone predicts a single N-terminal transmembrane domain with the majority of the protein exposed to the periplasm [86 , 87] . The periplasmic exposed C-terminal domain of CpaB was predicted by DELTA BLASTp to contain a domain from the SRPBCC superfamily present in the co-chaperone eukaryotic protein Aha1 , the activator of Hsp90 complex [56] . The SRPBCC domains are predicted to have deep hydrophobic ligand binding pockets . A BLASTp search of CpaB orthologs outside of Acinetobacter only identified two weak orthologs from Lysobacter antibioticus; however , a DELTA BLASTp search for CpaB orthologs outside of Acinetobacter primarily identified Aha1 as the closet ortholog , suggesting a possible eukaryotic ancestry . Currently , we hypothesize that the CpaA metallopeptidase is trafficked through the Sec system , as is the case for most type II secreted substrates . There , CpaA can interact with CpaB as CpaB is predicted to contain a single transmembrane domain with the majority of the protein exposed to the periplasmic space . Upon entry into the periplasmic space of CpaA from the Sec system , CpaB could facilitate proper folding of CpaA due to the requirement of type II secretion systems for competently folded proteins for active secretion . The potential role of the CpaA metallopeptidase in Acinetobacter pathogenesis and evolution is quite intriguing . Firstly , the type strains A . baumannii ATCC 17978 and 19606 , two of the more primitive Acinetobacter spp . used as model organisms do not contain orthologs of the CpaAB system , indicating a horizontal acquisition event within the last 70 years . Analysis of the GC content of the cpaAB locus and the surrounding DNA support this hypothesis . It is tempting to speculate , that given the predicted recent acquisition of the CpaAB protease/chaperone system and the role of the T2SS in Acinetobacter virulence , CpaA may be one of the major virulence factors of some pathogenic Acinetobacter spp . Future work will be aimed at deciphering the role of CpaA in the virulence assays utilized within this study . As mentioned above , LipB and CpaB act as specific chaperones for LipA and CpaA respectively . Some effectors secreted via a type III secretion system ( T3SS ) also require specific chaperones that have collectively been named “T3SS chaperones” [88] . T3SS chaperones do not present sequence similarity , but they are easily identified because they are encoded next to their cognate effector and most of them contain similar molecular weight and isoelectric points . Similarly , we define a “T2SS chaperone” as a protein encoded adjacently and co-regulated with a type II effector , that contains both an N-terminal transmembrane domain , and an exposed C-terminal region to the periplasm , and that is required for secretion of the cognate effectors . We identified “type II chaperones” in multiple Gram-negative species . Interestingly , LipB from Pseudomonas aeruginosa is a previously characterized chaperone that serves two T2SS effectors , LipA which is encoded next to LipB as well as LipC , which is encoded more than 2 Mb away [72] . This indicates that the T2SS chaperones family may be more widespread than we propose here . We determined that the Acinetobacter T2SS was required for virulence . We first determined that the mutants unable to produce a functioning T2SS were attenuated in the G . mellonella infection model . Given the high level of concordance between mutants attenuated in the G . mellonella model and mammalian models , we hypothesized a more relevant in vivo role for the Acinetobacter T2SS . We thus choose to investigate the role of T2S in a murine pulmonary infection model . Specifically , we observed high CFUs for the wild-type strain in the lungs after 36h infection period and also observed dissemination to both the liver and spleen . Using an unmarked , in-frame deletion of gspD strain and its respective complemented strain , we were able to demonstrate that the T2SS was indeed required for optimal colonization of both the lungs and spleen , but not the liver . Remarkably , we observed almost a two log decrease in CFUs in the lungs and spleen of mice infected with the gspD mutant strain when compared to either the wild type or the complemented strain . Many studies focusing on Acinetobacter pathobiology have utilized a similar murine pneumonia model of infection and also observed differences of around 2 logs; however , these mutants had defects in two-component regulatory systems , metabolism , and/or stress responses , all of which could have more pronounced global effects on Acinetobacter biology that mediate defects in colonization [32 , 89] . Herein , we have provided evidence of both a functional T2SS in many Acinetobacter spp . as well as demonstrated its importance in Acinetobacter pathogenicity . However , the exact role for each T2S effector proteins in Acinetobacter pathogenicity has yet to be determined . As such we plan to next probe the role of specific effectors in mediating the colonization phenotypes observed , with an emphasis on the most highly secreted protein , the CpaA metallopeptidase . Furthermore , our study highlights the use of other clinically relevant members of the genus Acinetobacter outside of A . baumannii in order to gain insights into the pathogenesis of clinically relevant Acb members . Although type strains like A . baumannii ATCC 17978 and 19606 have served well as model strains for Acinetobacter pathogenicity , their relative old age makes them less representative of current epidemic strains , which contain more antibiotic resistance cassettes and possibly novel virulence attributes .
Bacterial strains and plasmids utilized within this study can be located in the S1 Table . All bacterial strains were grown on L-agar at 37°C . Antibiotic selection for E . coli strains was used at the following concentrations: 100μg ampicillin/mL , 5μg tetracycline/mL , and 20μg kanamycin/mL . Antibiotic selection for Acinetobacter strains was used at the following concentrations: 200μg ampicillin/mL , 5μg tetracycline/mL , 20μg kanamycin/mL , 12 . 5μg chloramphenicol/mL . Sucrose was used at a final concentration of 10% for counter selecting Acinetobacter strains that lost the sacB cassette . All marked and unmarked mutants were generated using the previously published methodologies found in [22 , 45] using the In-Fusion HD EcoDry cloning kit . The In-Fusion HD EcoDry cloning kit was used to generate the interrupted gene constructs as described in the supplemental material of [22] and introduced into strain M2 via natural transformation as described in [45] . For strains containing the kan-sacB cassette , a tri-parental mating strategy was used to transiently introduce the pFLP2 plasmid as described in [23] , in order to replace the cassette with an frt scar . Strains designated with the “::frt” nomenclature contain a frt scar in place of the target gene . Each mutation was complemented using the mTn7 described in [22] . A complete list of primers for mutational analyses can be found in S2 Table . The Basic Local Alignment Search Tool ( BLAST ) tool was utilized in order to identify known gene homologs of type II secretion system related genes in Acinetobacter . Fifty milliliter cultures of each strain was grown for 18 h in M9 salts supplemented with 1% casamino acids and 1% glucose with 180 rpm . The secreted proteins were separated from the whole cells by centrifugation at 4000rpm for 10 mins . The supernatants were then further purified by filtration through 0 . 22 micron filters . The secreted proteins were then concentrated to ~100μL using Amicon Ultra Centrifugal Filter units with a 10kDa cutoff . Laemmli buffer with β-mercaptoethanol was added to each fraction and the samples were heated to 100°C by boiling in water for 10 mins . Twenty microliters of each sample was then separated by SDS-PAGE in a 4–20% gradient gel and subsequently silver stained . Secreted proteins used for the 2D-DIGE analysis were prepared as described in the above section discussing 1D SDS-PAGE analysis of secreted proteins for both the wild type A . nosocomialis strain M2 and its isogenic gspD::kan mutant . A detailed protocol for the 2D-DIGE analysis can be located in S1 Appendix . All 2D-DIGE experiments were performed by the Campus Chemical Instrument Center Mass Spectrometry and Proteomics Facility at The Ohio State University . In order to validate the 2D-DIGE analysis identifying the putative type II secreted proteins of strain M2 , selected effectors and effector/chaperone pairs were cloned into the Acinetobacter-E . coli shuttle vector pWH1266 . Briefly , lipA , cpaA , lipH , lipBA , and cpaAB loci were PCR amplified using the primers listed in S2 Table using strain M2 genomic DNA as template for PCRs . Each PCR product was purified , digested with PvuI-HF , and ligated into pWH1266 that was predigested with PvuI-HF and treated with phosphatase . The ligations were transformed into E . coli TOP10 cells and transformants were selected for on L-agar supplemented with tetracycline . Transformants were sub-cultured and each plasmid was purified and verified by sequencing . The carboxy-terminal His tag was added to lipA , lipH , and cpaA with a second PCR , where the respective forward primer included a 5’ overhang encoding for the His-tag using with the primers listed in S2 Table . The PCR products were purified , DpnI treated , and self-ligated . The ligations were transformed into TOP10 cells and transformants were selected on L-agar supplemented with tetracycline . Transformants were sub-cultured and the plasmids were purified and verified by sequencing . Vectors expressing the His-tagged constructs were electroporated into electrocompetent Acinetobacter spp . and transformants were selected for on L-agar supplemented with tetracycline . To test for PilD-dependent processing of GspG , the gspFG locus including the predicted native promoter was PCR amplified using the primers listed in S2 Table . The PCR product was purified , digested with PvuI-HF , and ligated into pWH1266 that was predigested with PvuI-HF and treated with phosphatase . The ligations were transformed into TOP10 cells and transformants were selected on L-agar supplemented with tetracycline . Transformants were sub-cultured and the plasmids were purified and verified by sequencing . To remove the gspF gene , an inverse PCR strategy was employed to PCR out gspF leaving the ATG start codon and the last 21 bp in order to generate an in-frame deletion . The PCR product was purified , treated with kinase , and self-ligated . The ligations were transformed into TOP10 cells and transformants were selected on L-agar supplemented with tetracycline . Transformants were sub-cultured and the plasmids were purified and verified by sequencing . The FLAG tag was PCR amplified onto the carboxy terminus of gspG as described above using the primers listed in S2 Table . The PCR product was purified , treated with kinase , and self-ligated . The ligations were transformed into TOP10 cells and transformants were selected on L-agar supplemented with tetracycline . Transformants were sub-cultured and the plasmids were purified and verified by sequencing . The pWH-gspG-FLAG construct was then electroporated into electrocompetent A . nosocomialis strains . Strains carrying His-tagged lipA , lipH , or cpaA were screened for active secretion via immunoblotting . Briefly , strains were struck and grown overnight on L-agar supplemented with tetracycline at 37°C . Bacteria were swabbed from the plate , resuspended in LB broth , and used to inoculate 10mL of LB broth to an OD600 of 0 . 05 supplemented with tetracycline . The cultures were grown to mid-log phase , normalized to an OD600 of 0 . 5 , then processed for whole cell fractions and secreted fractions . Whole cell fractions were obtained by removing 1mL of the normalized mid-log cells , pelleting the cells by centrifugation , and removing the supernatant . Bacterial pellets were then resuspended in 50μL of 1X Laemmli buffer . Secreted fractions were obtained by pelleting the normalized mid-log cultures by centrifugation and carefully removing 1mL of the supernatant . Secreted proteins were precipitated by the addition of 250μL of a saturated trichloroacetic acid solution . Precipitated proteins were incubated on ice for 10–30 mins , pelleted by centrifugation , and washed twice with ice-cold acetone . Residual acetone was removed by heating the samples at 95°C . Precipitated proteins were resuspended in 100μL of 1X Laemmli buffer . Both whole cell fractions and secreted fractions were boiled in 1X Laemmli buffer for 10 mins and subsequently used for immunoblotting . Proteins were separated on a 10% SDS-PAGE gel , transferred to nitrocellulose , and probed for RNA polymerase and 6X-Histidine tagged proteins according to our previously published methodologies . In order to determine lipolytic activity of secreted protein fractions , a modified version of the para-nitrophenol palmitate ( p-NPP ) lipase assay was performed . Secreted protein fractions were purified from select strains as described above with slight modifications . Briefly , 2 . 5mL of culture supernatant was clarified by centrifugation and then filtered through 0 . 22μM PVDF filters . The secreted proteins were buffer exchanged into 50mM Tris and were concentrated to ~250μL using Amicon Ultra Centrifugal Filter units with a 10kDa cutoff and promptly used for the lipase assay . Lipase activity was determined by measuring the absorbance at 410nm at 37°C using p-NPP as a substrate . The p-NPP solution was freshly prepared for each assay by diluting solution A ( 0 . 1g p-NPP in 100mL isopropanol ) 1:10 with solution B ( 1g gum Arabic , 2g sodium deoxycholate , 5mL triton X-100 , 50mM Tris-HCl pH 8 in 900mL ) . Seventy microliters of the p-NPP solution was then added to 30μL of the concentrated , clarified secreted protein fractions from a respective strain . Kinetic measurements recording the absorbance at 410nm were then performed over the designated time frame at 37°C with orbital shaking between each absorbance reading . Absorbance measurements were captured using the Synergy HTX multi-mode reader from BioTek . Each experiment was performed in triplicate with three technical replicates per sample . A . nosocomialis M2 , the ΔgspD::kan mutant and the complemented strain were grown in LB broth overnight in an orbital shaker ( 37°C , 200rpm ) . The overnight cultures were diluted to a starting OD600 0 . 05 and grown at 37°C with 200rpm to a final OD600 of 0 . 5 . 0 . 5ODs was pelleted by centrifugation , washed with filter sterilized PBS and resuspended at and OD of 0 . 5/ml , 0 . 158OD/mL and 0 . 05OD/mL in filter sterile PBS . The CFU/mL at 0 . 5OD/mL was determined to be 109 . Serial dilution of the 0 . 5OD/mL sample was performed . Larvae were injected with 10uL of sterile PBS , 106 or 107 CFU . 3 groups of 10 larvae were injected per experimental group . The larvae were scored as live/dead depending on their response to physical stimulus approximately every 5 hours . The number of bacterial cells injected into the larvae was determined by plating 10-fold serial dilutions on LB agar and performing CFU counts after overnight incubation at 37°C . All infection experiments were approved by the Vanderbilt University Institutional Animal Care and Use Committee . Wild-type C57BL/6 mice , obtained from Jackson Laboratories , were used for single infection experiments with either the wild type A . nosocomialis M2 , the M2∆gspD::frt mutant , or the respective gspD complemented strain . Overnight cultures of each strain were sub-cultured 1/1000 into 50 mL LB broth and grown with shaking at 37°C in 250-mL flasks . Bacterial cells were harvested by centrifugation during logarithmic growth , washed twice with phosphate buffered saline ( PBS ) , and suspended in PBS . Nine-week old male mice were inoculated intranasally with a total of 7–8 X 108 cfu in 30 μL . At 36 h post-infection , mice were euthanized and CFUs were enumerated from the lungs , livers , and spleens following tissue homogenization and dilution plating to LB agar medium . The data were log transformed and analyzed for Gaussian distribution using the D’Angostino-Pearson omnibus normality test . Data sets displaying Gaussian distribution were then analyzed by One-way ANOVA with Tukey’s test for multiple comparisons . Data sets displaying non-Gaussian distribution were analyzed by Kruskal-Wallis test with Dunn’s test for multiple comparisons . All statistical analyses were performed using GraphPad Prism 6 ( GraphPad Software Inc . , La Jolla , CA ) . Animal care and experiments were performed in accordance with the NIH “Guide for the Care and Use of the Laboratory Animals” and were reviewed and approved by the Vanderbilt University Institutional Animal Care and Use Committee ( Protocol M/10/165 ) . Mice were anesthetized with 2 , 2 , 2 , -tribromoethanol prior to intranasal inoculation . Mice were euthanized by carbon dioxide . | Members of the genus Acinetobacter , specifically , A . baumannii , A . pittii , and A . nosocomialis , have rapidly emerged as opportunistic human pathogens particularly targeting the immunocompromised patient population . Of significant concern is the fact that many Acinetobacter-induced infections are caused by multiply-drug resistant strains severely limiting clinical intervention strategies . In order to best develop new therapeutic treatment options against Acinetobacter infections , we first must gain insight into the mechanisms these bacteria utilize to cause disease . One common way bacteria mediate pathogenesis is through the secretion of proteins or toxins . Here we identified and examined the role of a type II secretion system in Acinetobacter biology and pathogenesis . We found that A . nosocomialis secretes multiple proteins through the type II secretion system , including two lipases and a protease . Furthermore , two of the secreted proteins required dedicated membrane-associated chaperones for secretion . These chaperones appear to be present in multiple bacterial species . Lastly , we found that the Acinetobacter type II secretion system was required for full virulence in a murine pulmonary infection model , indicating that this secretion system may be used during the course of an Acinetobacter infection . Collectively , we have uncovered a new mechanism by which Acinetobacter species mediate disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| []
| 2016 | Medically Relevant Acinetobacter Species Require a Type II Secretion System and Specific Membrane-Associated Chaperones for the Export of Multiple Substrates and Full Virulence |
The plant hormone auxin regulates numerous growth and developmental processes throughout the plant life cycle . One major function of auxin in plant growth and development is the regulation of cell expansion . Our previous studies have shown that SMALL AUXIN UP RNA ( SAUR ) proteins promote auxin-induced cell expansion via an acid growth mechanism . These proteins inhibit the PP2C . D family phosphatases to activate plasma membrane ( PM ) H+-ATPases and thereby promote cell expansion . However , the functions of individual PP2C . D phosphatases are poorly understood . Here , we investigated PP2C . D-mediated control of cell expansion and other aspects of plant growth and development . The nine PP2C . D family members exhibit distinct subcellular localization patterns . Our genetic findings demonstrate that the three plasma membrane-localized members , PP2C . D2 , PP2C . D5 , and PP2C . D6 , are the major regulators of cell expansion . These phosphatases physically interact with SAUR19 and PM H+-ATPases , and inhibit cell expansion by dephosphorylating the penultimate threonine of PM H+-ATPases . PP2C . D genes are broadly expressed and are crucial for diverse plant growth and developmental processes , including apical hook development , phototropism , and organ growth . GFP-SAUR19 overexpression suppresses the growth defects conferred by PP2C . D5 overexpression , indicating that SAUR proteins antagonize the growth inhibition conferred by the plasma membrane-localized PP2C . D phosphatases . Auxin and high temperature upregulate the expression of some PP2C . D family members , which may provide an additional layer of regulation to prevent plant overgrowth . Our findings provide novel insights into auxin-induced cell expansion , and provide crucial loss-of-function genetic support for SAUR-PP2C . D regulatory modules controlling key aspects of plant growth .
The plant hormone auxin regulates nearly all aspects of growth and development , including embryogenesis [1] , root development [2] , gravitropism [3] , leaf development [4] , vascular development [5] , phototropism [6] , shade avoidance [6] , shoot apical meristem development [7] , flower primordium formation [5] , stamen development [8] , and gynoecium development [9] . At the cellular level , auxin regulates these processes through the control of cell division , expansion , and differentiation [10 , 11] . Auxin is perceived by a co-receptor complex that is composed of TRANSPORT INHIBITOR RESPONSE 1/AUXIN SIGNALING F-BOX PROTEINS ( TIR1/AFBs ) and AUXIN/INDOLE-3-ACETIC ACID ( AUX/IAA ) transcriptional repressors in the nucleus . Auxin binding of the TIR1/AFB and AUX/IAA complex leads to the degradation of AUX/IAA proteins by the 26S proteasome [12] . AUX/IAA degradation then relieves the repression of AUXIN RESPONSE FACTOR ( ARF ) transcription factors to activate the expression of auxin-responsive genes [13] . These auxin-responsive genes , including SMALL AUXIN UP RNAs ( SAURs ) [14] , then regulate auxin-mediated cellular , physiological , and developmental processes . A major function of auxin in plant growth and development is the regulation of cell expansion . Auxin-induced cell expansion was hypothesized to occur via an acid growth mechanism , which was first proposed in the 1970s [15 , 16] . According to this theory , auxin activates plasma membrane ( PM ) H+-ATPases ( known as AHAs/ARABIDOPSIS H+-ATPases in Arabidopsis ) , which pump protons across the plasma membrane , thus acidifying the apoplast and elevating membrane potential . The more acidic apoplastic pH activates expansins and other cell wall remodeling enzymes , resulting in increased cell wall extensibility . Additionally , plasma membrane hyperpolarization promotes increased solute and water uptake , providing elevated turgor to drive cell expansion . In recent years , several studies have provided molecular support for auxin-mediated PM H+-ATPase activation during hypocotyl growth in Arabidopsis . Auxin induces the phosphorylation of the penultimate threonine ( Thr947 in AHA2 ) , a key regulatory site of PM H+-ATPases without altering AHA protein abundance [17] . This increase in AHA Thr947 phosphorylation is likely the result of auxin-induced SAUR expression , as GFP-SAUR19 overexpression promotes AHA-Thr947 phosphorylation , hypocotyl elongation , and cell wall extensibility [18 , 19] . Furthermore , consistent with the hypothesis that acid growth requires auxin-mediated gene expression , the canonical TIR1/AFB-AUX/IAA nuclear signaling pathway was recently shown to be required for auxin-induced hypocotyl elongation and cell wall acidification [20 , 21] . Together , these findings provide strong genetic and biochemical evidence in support of the acid growth theory . SAUR proteins promote PM H+-ATPase activation and the resulting cell expansion by inhibiting the activity of type 2C protein phosphatases belonging to the D subfamily ( PP2C . D ) [18 , 19] . PP2Cs are Mg2+/Mn2+-dependent enzymes that are evolutionarily conserved from prokaryotes to eukaryotes [22 , 23] . The Arabidopsis genome encodes eighty PP2Cs , nine of which belong to the D-subclade [22] . Recent studies have implicated PP2C . D family members in the regulation of apical hook development [18 , 24] , auxin-induced cell expansion [18] , leaf senescence [25] , and immune response [26] . In our previous work [18] , we found that plants harboring an artificial microRNA ( amiRNA ) targeting five D-clade family members conferred phenotypes similar to , albeit much weaker than , GFP-SAUR19 overexpression lines , including increased hypocotyl length , hypersensitivity to LiCl , and increased medium acidification . While this finding provided initial genetic support for an antagonistic role of SAUR and PP2C . D proteins in regulating PM H+-ATPase activity and cell expansion , the identities and functional relationships of the specific PP2C . D family members involved remained uncertain . Furthermore , the amiRNA reverse genetic approach does not completely abolish gene function and can be prone to off-targeting effects . Here , we extend our studies on the functions of these phosphatases by conducting a genetic characterization of pp2c . d family loss-of function mutants . We find that the PP2C . D family members exhibit distinct subcellular localization patterns , and the plasma membrane-localized subset , PP2C . D2 , D5 , and D6 phosphatases , play a major role in antagonizing SAUR-mediated regulation of PM H+-ATPase activity , cell expansion , and plant growth and development .
To investigate the roles of the PP2C . D family genes in plant growth and development , we examined PP2C . D gene expression using the GUS ( β-glucuronidase ) reporter gene driven by the native PP2C . D promoters . We generated Arabidopsis transgenic plants expressing PP2C . D1pro:EGFP-GUS transcriptional or PP2C . D ( 2–9 ) pro:PP2C . D ( 2–9 ) -GUS translational reporter constructs . All PP2C . D genes except PP2C . D7 were expressed in the cotyledons and hypocotyls of light-grown seedlings ( Fig 1A ) . In the roots of light-grown seedlings , all PP2C . D genes except PP2C . D1 and PP2C . D7 were ubiquitously expressed ( Fig 1B ) . PP2C . D1 was weakly expressed specifically in the root elongation zone , while PP2C . D7 was weakly expressed throughout the root except for the root tip region ( Fig 1B ) . All PP2C . D genes were also expressed in the cotyledons and hypocotyls of etiolated seedlings except for PP2C . D7 , which was only expressed in the hypocotyls ( Fig 1C ) . Interestingly , PP2C . D1 exhibited stronger expression on the inner side of the apical hook ( Fig 1C and 1D ) , consistent with previous findings implicating PP2C . D1 in apical hook formation [18 , 24] . All PP2C . D genes except PP2C . D7 were expressed in the petioles and rosette leaves of 2-week-old plants , while PP2C . D7 was only very weakly expressed in the petioles ( S1A Fig ) . In flowers and siliques , all PP2C . D genes except PP2C . D7 were expressed in multiple floral organs and siliques , while PP2C . D7 was only weakly expressed in petals ( S1B and S1C Fig ) . In particular , PP2C . D1 , D2 , D5 , and D6 were strongly expressed in stamen filaments . PP2C . D8 expression was highly enriched in pistils , and PP2C . D4 and D6 exhibited strong expression in anthers . The absence or low level of PP2C . D7 expression in nearly all organs examined is consistent with the results of numerous transcriptomic studies compiled in the Arabidopsis eFP Browser database ( S2 Fig ) [27] . Together , our GUS expression data demonstrate that all PP2C . D genes except PP2C . D7 are broadly expressed in most organs and tissues during the life cycle of plants , especially in the growing organs and tissues , including hypocotyls , young roots and leaves , and stamen filaments . These results indicate that the PP2C . D family genes may play an important role in a variety of plant growth and developmental processes . The subcellular localization of the PP2C . D family phosphatases was previously examined using green fluorescent protein ( GFP ) as a reporter in Arabidopsis transgenic plants harboring reporter constructs driven by the cauliflower mosaic virus ( CaMV ) 35S promoter [28] . However , since overexpression driven by the 35S promoter may cause fluorescent fusion protein mislocalization , we further examined the subcellular localization of the PP2C . D family phosphatases using the native PP2C . D promoters to express PP2C . D-GFP fusion proteins . We generated PP2C . D ( 1–9 ) pro:PP2C . D ( 1–9 ) -GFP Arabidopsis transgenic plants . Since our GUS reporter analysis indicated that all PP2C . D family members except PP2C . D1 and D7 were strongly expressed in root tips ( Fig 1B ) , we initially examined PP2C . D-GFP localization in cells of the root meristem and elongation zone . Previous studies reported that PP2C . D1 , also known as APD7 ( Arabidopsis PP2C clade D 7 ) [28] or SSPP ( SENESCENCE-SUPPRESSED PROTEIN PHOSPHATASE ) [25] , resided in the nucleus and cytoplasm of root cells [28] or only in the cytoplasm of mesophyll protoplasts [25] . Consistent with our PP2C . D1pro:EGFP-GUS reporter findings , PP2C . D1-GFP fluorescence in root tips was below our detection limit . However , since PP2C . D1 is auxin-inducible [29] , when PP2C . D1pro:PP2C . D1-GFP seedlings were treated with IAA , robust expression was observed , with GFP fluorescence evident in both nuclei and the cytosol ( Fig 2A ) . To better examine PP2C . D1 localization without the complications of exogenous IAA , we examined apical hooks of etiolated seedlings . Consistent with the GUS expression pattern ( Fig 1C and 1D ) , GFP fluorescence was specifically observed in the epidermal layer of the inner side of the hook ( Fig 2B ) . Similar to auxin-treated roots , PP2C . D1 appeared to localize to both nuclei and the cytoplasm . Consistent with previous findings using 35S-driven GFP reporters [28] , PP2C . D2/APD2 , PP2C . D5/APD6 , and PP2C . D6/APD3 localized exclusively to the cell periphery of root cells , consistent with a plasma membrane localization ( Fig 2A ) . PP2C . D3/PP2C38 was recently reported to reside at the plasma membrane and in intracellular punctae of Arabidopsis leaf epidermal cells overexpressing PP2C . D3-GFP driven by the 35S promoter [26] . In contrast to these results , while we did detect some PP2C . D3-GFP signal near the cell periphery , considerable fluorescence was also observed in intracellular punctae and nuclei ( Fig 2A ) . Likewise , PP2C . D4/APD4 exhibited both nuclear and cytosolic localization ( Fig 2A ) . PP2C . D7/APD9 was reported to associate with the plasma membrane and endomembranes [28] . Although we examined many independent PP2C . D7pro:PP2C . D7-GFP lines , we could not successfully detect PP2C . D7-GFP protein expression in any organs examined , consistent with the low expression observed with the GUS reporter ( Fig 1B ) and transcriptomic studies ( S2 Fig ) . PP2C . D8-GFP and MitoTracker signals colocalized in root cells ( Fig 2C ) , supporting the previous findings that PP2C . D8/APD5 is localized in the mitochondria [28] . Lastly , Tovar-Mendez et al . ( 2014 ) reported that PP2C . D9/APD8 localized to the cytoplasm , but not in the nucleus [28] . Our results support this conclusion , although similar to PP2C . D3 , cytosolic fluorescence was frequently punctate rather than uniform ( Fig 2A ) . While we cannot be certain that all PP2C . D-GFP proteins are functional and localize precisely like the endogenous protein , we provide evidence below that the PP2C . D2- , D5- , and D6-GFP constructs encode functional proteins ( S5C Fig ) . Given the high degree of sequence similarity between PP2C . D family members , it seems likely that the addition of a C-terminal GFP tag does not interfere with PP2C . D function . Together , our results indicate that the PP2C . D family phosphatases exhibit distinct subcellular localization , residing in various cellular compartments . The Arabidopsis hypocotyl is an excellent system to study elongation growth , since its size is mainly controlled by cell expansion [30] . Our previous studies using an amiRNA to knock down multiple members ( PP2C . D2 , D5 , D7 , D8 , D9 ) of the PP2C . D family genes showed that the PP2C . D family phosphatases may function redundantly to negatively regulate hypocotyl growth [18] . The resulting growth phenotypes , however , were weak in comparison to GFP-SAUR19 overexpression plants . To definitively determine the contribution of individual PP2C . D phosphatases to cell expansion , we analyzed the hypocotyl growth phenotype of all pp2c . d T-DNA insertion mutants ( S3 Fig ) . Semiquantitative RT-PCR analysis confirmed that the pp2c . d1 , d2 , d5 , d6 , d7 , d8 , and d9 insertion mutants were likely null alleles ( S4 Fig ) . Likewise , the pp2c . d3 and d4 insertion mutants were previously reported to be null or severe knockdown mutants [26] . pp2c . d5 was the only single mutant that exhibited slightly increased hypocotyl growth ( S5A Fig ) . The pp2c . d4 T-DNA insertion mutant was unavailable when we started to generate various pp2c . d higher order mutants . The pp2c . d3/4 double mutant was recently published [26] , and the mutant seedlings did not exhibit an obvious hypocotyl growth phenotype ( S5B Fig ) . The lack of strong single mutant phenotypes suggested functional redundancy within the PP2C . D gene family . We therefore generated a variety of double , triple , and quadruple mutants ( S5A Fig ) . Given the plasma membrane localization of SAUR19 and PM H+-ATPases , we were particularly interested in lines that lack the three plasma membrane-localized family members , PP2C . D2 , D5 , and D6 . Interestingly , mutations in PP2C . D2 or PP2C . D6 enhanced the hypocotyl growth phenotype of pp2c . d5 ( S5A Fig ) . Likewise , the pp2c . d2 pp2c . d6 double mutant exhibited a long hypocotyl phenotype . Hypocotyl length was unaffected , however , in all of the other double mutant combinations tested . Furthermore , the pp2c . d2/5/6 triple mutant seedlings displayed an even stronger hypocotyl growth phenotype , exhibiting hypocotyls nearly as long as those of GFP-SAUR19 overexpression seedlings ( Figs 3A , 3B and S5A ) . Like GFP-SAUR19 seedlings , this increase in hypocotyl length was the result of increased cell expansion ( Fig 3C ) . In contrast , various combinations of triple mutants for pp2c . d1 , d3 , d8 , and d9 did not exhibit any obvious hypocotyl growth phenotype ( S5A Fig ) . We previously found that PP2C . D1 overexpression under the control of the 35S promoter conferred a dramatic reduction in hypocotyl length and plant stature [18] . Loss-of-function analysis , however , indicates that endogenous PP2C . D1 plays little if any role in hypocotyl elongation under our growth conditions , as the pp2c . d1 mutation failed to enhance the hypocotyl growth phenotype of the pp2c . d2/5/6 triple mutant as well as lower order mutant combinations ( S5A Fig ) . To confirm that the pp2c . d2/5/6 long hypocotyl phenotype was in fact due to loss of the three PM-localized phosphatases , we transformed the triple mutant with the PP2C . D ( 2 , 5 , 6 ) pro:PP2C . D ( 2 , 5 , 6 ) -GFP reporter constructs used to assess localization ( Fig 2 ) . All three GFP fusion constructs restored hypocotyl length to at least the corresponding double mutant ( S5C Fig ) . In the D2- and D5-GFP lines , over-complementation was observed , with hypocotyl lengths returning wild-type length . Presumably , this is due to position effects that may result in modest PP2C . D overexpression . Together , the above genetic findings indicate that the plasma membrane-localized PP2C . D2 , D5 , and D6 proteins are the major PP2C . D phosphatases that negatively regulate cell expansion during hypocotyl growth . Since the pp2c . d2/5/6 triple mutant exhibited a long hypocotyl phenotype , we proceeded to assess this mutant for other GFP-SAUR19 overexpression-related phenotypes . Like GFP-SAUR19 seedlings , pp2c . d2/5/6 seedlings exhibited dramatic hypersensitivity to 10 mM LiCl ( Fig 3D ) and increased medium acidification ( Fig 3E ) , phenotypes suggestive of elevated PM H+-ATPase activity [18 , 31] . To test this possibility , we examined AHA-Thr947 phosphorylation status indirectly using a GST-14-3-3 far western blotting assay . Several prior studies have demonstrated that this assay accurately reflects AHA-Thr947 phosphorylation status and the corresponding changes in PM H+-ATPase activity [17 , 18 , 32 , 33] . A striking increase in AHA-Thr947 phosphorylation was observed in both GFP-SAUR19 and pp2c . d2/5/6 seedlings ( Fig 3F ) . In contrast , AHA-Thr947 phosphorylation levels in the pp2c . d1/3/8 triple mutant were not noticeably different from wild-type ( Fig 3F ) . The strong genetic interaction observed in hypocotyl growth assays suggested that PP2C . D2 , D5 , and D6 act in a redundant fashion ( S5A Fig ) . Consistent with this notion , while the pp2c . d2/5/6 triple mutant exhibited increased AHA-Thr947 phosphorylation levels , phosphorylation in the single mutants was comparable to wild-type ( Fig 3F ) . Together , our genetic and biochemical findings demonstrate that these PM-localized PP2C . D family members function redundantly to regulate PM H+-ATPase activity to control cell expansion . Although pp2c . d2/5/6 plants exhibited no major developmental abnormalities , several growth phenotypes were apparent in older plants . Three-week-old pp2c . d2/5/6 plants exhibited slightly larger rosette leaves than those of wild-type , but similar to the case for hypocotyl growth , this phenotype was slightly more dramatic in GFP-SAUR19 plants ( Fig 3G and 3H ) . In flowers , pp2c . d2/5/6 flowers exhibited longer stamen filaments and pistils than those of wild-type flowers ( Fig 3I ) . While GFP-SAUR19 flowers did not exhibit long stamen filament and pistil phenotypes ( Fig 3I ) , increased stamen filament length has been reported for plants expressing SAUR63-GFP or GUS fusion proteins [34] . Plant height and silique length of mature pp2c . d2/5/6 plants were also slightly larger than wild-type ( S5D and S5E Fig ) . While GFP-SAUR19 plants did not exhibit an obvious silique growth phenotype ( S5E Fig ) , increased silique growth was recently reported in transgenic Arabidopsis plants overexpressing SAUR8 , SAUR10 , and SAUR16 [35] . PP2C . D1 is differentially expressed in the apical hook of etiolated seedlings ( Figs 1D and 2B ) , and etiolated pp2c . d1 mutants , as well as GFP-SAUR19 seedlings , exhibit defective apical hook development [18 , 24] . While the pp2c . d2 , d5 , and d6 single mutants develop apical hooks comparable to wild-type [18] , given the functional redundancy we observed in other assays of these family members , we examined the apical hook phenotype of the pp2c . d2/5/6 triple mutant . Indeed , like etiolated GFP-SAUR19 seedlings , pp2c . d2/5/6 seedlings exhibited partially opened apical hooks and expanded cotyledons ( Fig 3J and 3K ) . In our previous work [36] , GFP-SAUR19 seedlings were shown to exhibit reduced phototropism , suggesting the involvement of SAUR proteins in tropic growth responses . Consistent with this notion , SAUR transcripts have been found to preferentially accumulate on the elongating side of bending organs [37–39] . We therefore examined whether PP2C . D2 , D5 , and D6 may function in phototropic response . When exposed to unilateral blue light , etiolated pp2c . d2/5/6 seedlings exhibited dramatically reduced phototropic curvature ( Fig 3L ) , suggesting that SAUR-mediated inhibition of PP2C . D2/5/6 activity on the light distal side of the hypocotyl may underlie phototropic bending . Based on the increased growth phenotypes of pp2c . d2/5/6 plants and their strong phenotypic similarity to SAUR gain-of-function plants , as well as similar effects on PM H+-ATPase Thr947 phosphorylation , our results suggest that PP2C . D2 , D5 , and D6 phosphatases are the primary effectors of plasma membrane-localized SAUR proteins that regulate plant growth . The above genetic studies revealed that PP2C . D2 , D5 , and D6 are the major D-clade phosphatases that negatively regulate SAUR-mediated cell expansion . Our previous work isolated PP2C . D1 , D5 , and D6 as SAUR19 interacting proteins in a yeast two-hybrid library screen [18] . We confirmed that PP2C . D2 also interacted with SAUR19 in this system ( Fig 4A ) . We also tested the remaining PP2C . D family members and found that PP2C . D3 , D4 , and D8 can also interact with SAUR19 ( S6A Fig ) . Positive interactions were not detected for PP2C . D7 or PP2C . D9 ( S6A Fig ) , however , PP2C . D7 did not appear to be expressed in yeast and PP2C . D9 expression was quite low in comparison to PP2C . D1 ( S6B Fig ) . To examine protein interactions in planta , we generated transgenic Arabidopsis plants co-expressing PP2C . D2-HA , D5-HA , or D6-HA under the control of native PP2C . D promoters and GFP-SAUR19 driven by the 35S promoter and examined their interactions by co-immunoprecipitation ( co-IP ) using solubilized microsomal fractions . PP2C . D5-HA and GFP-SAUR19 co-immunoprecipitated , confirming their interaction in Arabidopsis ( Fig 4B ) . However , we could not successfully co-immunoprecipitate GFP-SAUR19 with PP2C . D2-HA and PP2C . D6-HA . While this may be due to the technical limitations of this assay , we cannot exclude the possibility that SAUR19 does not interact with these phosphatases in planta . Rather , given the large number of SAUR proteins , it seems quite possible that PP2C . D2 and D6 may preferentially interact with other plasma membrane-associated SAUR proteins . The pronounced increase in AHA-Thr947 phosphorylation observed in the pp2c . d2/5/6 triple mutant ( Fig 3F ) identified this phosphosite as a putative substrate of PP2C . D2 , D5 , and D6 phosphatases . We therefore examined potential interactions between these proteins and PM H+-ATPases by co-IP and bimolecular fluorescence complementation ( BiFC ) assays . PP2C . D2-GFP , D5-GFP , and D6-HA all co-immunoprecipitated with AHA proteins ( Fig 4C ) . In contrast , no detectable interaction was observed between AHAs and PP2C . D8-GFP ( Fig 4C ) , suggesting at least some degree of substrate specificity among the PP2C . D family members . Additionally , yellow fluorescent signals were observed at the plasma membrane of leaf epidermal cells when AHA2-YFPN was transiently co-expressed with PP2C . D2-YFPC , D5-YFPC , and D6-YFPC in Nicotiana benthamiana leaves ( S7 Fig ) . These results indicate that PP2C . D2 , D5 , and D6 phosphatases physically associate with AHA proteins in planta . To further test the regulatory nature of these interactions , we co-expressed PP2C phosphatases with AHA2 in yeast strain RS-72 for complementation assays . In this strain , cells are only viable when grown on galactose media , since the endogenous yeast PM H+-ATPase gene PMA1 is driven by the GAL1 promoter . AHA2 expression complements GAL1pro:PMA1 to restore growth on glucose media [33 , 40] . When we co-expressed PP2C . D2 , D5 , or D6 with AHA2 , yeast RS-72 cells were unable to grow on glucose media ( Fig 4D ) , indicating that these phosphatases inhibit AHA2 function . In fact , all PP2C . D family members with the exception of PP2C . D8 were capable of antagonizing AHA2 function in yeast . In contrast , the non-D-clade Arabidopsis PP2C phosphatases PP2C . I1 ( At2g25070 , an I-clade PP2C ) and PP2C . F9 ( At1g43900 , an F-clade PP2C ) [41] failed to inhibit AHA2 function in this system ( Fig 4D ) , suggesting that the D-clade PP2Cs may be unique in their ability to inhibit PM H+-ATPase activity . Presumably , PP2C . D1 , D3 , D4 , and D9 , all of which displayed some degree of cytosolic localization in Arabidopsis , can access the cytosolic C-terminus of AHA2 when overexpressed in yeast . Our genetic ( S5 Fig ) and biochemical ( Fig 3F ) findings , however , suggest that these family members are not the primary regulators of AHA activity in planta . We previously developed an in vitro AHA2 dephosphorylation assay using plasma membranes prepared from yeast RS-72 cells expressing AHA2 to examine SAUR regulation of PP2C . D1-mediated dephosphorylation of AHA2-Thr947P [18] . This same assay was used to assess the ability of PP2C . D2 , D5 , and D6 phosphatases to dephosphorylate AHA2-Thr947P and the inhibition of any such activity by SAUR proteins . As previously shown for PP2C . D1 [18] , recombinant PP2C . D2 and PP2C . D5 catalyzed AHA2-Thr947 dephosphorylation and this activity was strongly inhibited by the addition of purified SAUR9 protein ( Fig 4E ) . We could not demonstrate phosphatase activity for recombinant PP2C . D6 in this system or in assays employing the chemical substrate p-nitrophenyl phosphate ( pNPP ) , suggesting that this phosphatase may require alternative reaction conditions , co-factors , or post-translational modifications . That said , we cannot eliminate the possibility that PP2C . D6 is not a functional phosphatase , and rather may play a distinct role such as providing a scaffolding function for PP2C . D-substrate complexes . However , given that PP2C . D6 contains a highly conserved catalytic domain , together with our interaction data and genetic findings demonstrating that D6 functions redundantly with D2 and D5 , it seems likely that PP2C . D6 also dephosphorylates AHA2-Thr947P and this activity is inhibited by SAUR proteins . To investigate the effects of PP2C . D gain-of-function on plant growth and development , we tried to generate 35Spro:PP2C . D5-EYFP overexpression lines . While several primary Arabidopsis transformants exhibited a dwarf phenotype , we could not obtain any stable homozygous overexpression lines , suggesting that PP2C . D5 dosage may be critical . We therefore generated transgenic Arabidopsis plants expressing PP2C . D5-HA driven by the PP2C . D5 promoter in the pp2c . d5 mutant background . PP2C . D5-HA protein expression rescued the slightly increased hypocotyl growth phenotype of pp2c . d5 seedlings ( S8A Fig ) , demonstrating that PP2C . D5-HA is a functional protein . We noticed that some transgenic lines exhibited growth defects , including reduced growth and fertility . The severity of growth defects was dependent on the expression levels of PP2C . D5-HA protein . pp2c . d5 PP2C . D5-HA lines 6 and 7 that did not exhibit obvious growth defects expressed lower levels of PP2C . D5-HA protein , while lines 1 and 4 that exhibited severe growth defects expressed higher levels of PP2C . D5-HA protein ( S8B Fig ) . We therefore selected pp2c . d5 PP2C . D5-HA lines 1 and 4 ( hereafter referred to as D5-HA-OX for Over-eXpression ) for further phenotypic analyses to assess the effects of PP2C . D5 gain-of-function on plant growth and development . Compared with wild-type and pp2c . d5 seedlings , light-grown D5-HA-OX seedlings exhibited reduced hypocotyl growth ( Fig 5A and 5B ) , shorter hypocotyl epidermal cells ( Fig 5C ) , and reduced root growth ( Fig 5D ) . Etiolated D5-HA-OX seedlings also exhibited severely reduced hypocotyl growth ( Fig 5E and 5F ) . D5-HA-OX plants exhibited smaller rosette leaves ( Fig 5G ) , delayed leaf senescence ( Fig 5H ) , and smaller flowers with shorter stamen filaments ( Fig 5J ) . Shortly after bolting , D5-HA-OX plants exhibited fertility defects ( Fig 5I ) . However , hand-pollination of D5-HA-OX pistils with D5-HA-OX pollen grains resulted in full seed set ( Fig 5K ) , indicating that the fertility defects are caused by reduced stamen filament elongation rather than defective pollen or fertilization . Curiously , with continued growth , older D5-HA-OX plants recovered from the early male fertility defects and could set seeds successfully . The mature D5-HA-OX plants were smaller than those of wild-type and pp2c . d5 ( Fig 5L ) and had shorter siliques ( Fig 5M ) . To assess whether PP2C . D5 overexpression confers reduced PM H+-ATPase activity , we examined the phosphorylation status of AHA-Thr947 in D5-HA-OX seedlings . A clear decrease in AHA-Thr947 phosphorylation was observed in etiolated D5-HA-OX seedlings ( Fig 5N ) . Consistent with a reduction in PM H+-ATPase activity and consequent membrane potential , D5-HA-OX seedlings exhibited resistance to 10 mM LiCl ( Fig 5O ) and reduced medium acidification ( Fig 5P ) . Together , our results suggest that PP2C . D5 overexpression confers reduced cell expansion and plant growth , which are caused at least in part by reduced PM H+-ATPase Thr947 phosphorylation and the corresponding reduction in enzyme activity . The Arabidopsis genome contains 79 SAUR genes . Due to extensive functional redundancy , it is challenging to study the functions of SAUR proteins in plant growth and development using a loss-of-function approach [14] . Our prior biochemical studies have demonstrated that SAUR proteins inhibit PP2C . D phosphatase activity ( Fig 4E ) [18] . To test this hypothesis genetically , we generated Arabidopsis plants co-expressing GFP-SAUR19 and PP2C . D5-HA proteins by crossing the 35Spro:GFP-SAUR19 transgene into D5-HA-OX line 1 , and examined the effects of GFP-SAUR19 overexpression on the growth defects of PP2C . D5-HA overexpression plants . Western blot analysis confirmed that the double transgenic plants expressed both fusion proteins at levels comparable to that seen in the parental lines ( Fig 6A ) . Strikingly , GFP-SAUR19 overexpression suppressed virtually all aspects of the D5-HA-OX phenotypes , including the hypocotyl , root , and rosette leaf growth defects ( Fig 6B–6E ) . In addition , the male sterility of D5-HA-OX plants caused by defective stamen filament elongation growth was also suppressed by GFP-SAUR19 overexpression ( Fig 6F and 6G ) , as were the D5-HA-OX defects in plant height and silique length ( Fig 6H and 6I ) . Together , these results provide strong genetic support for the hypothesis that SAUR proteins and plasma membrane-localized PP2C . D phosphatases function antagonistically to regulate plant growth . Previous hypocotyl transcriptomic analyses of auxin-responsive genes revealed that PP2C . D1 and PP2C . D7 may be auxin-induced genes , as their expression was upregulated by a 120 min treatment with the synthetic auxin picloram [42] . Additionally , using our PP2C . D1-GFP reporter , we demonstrated auxin-inducible expression of PP2C . D1 in root tips ( Fig 2A ) . To examine potential auxin-mediated regulation of PP2C . D family genes , we examined PP2C . D-GUS expression in 5-day-old light-grown PP2C . D1pro:EGFP-GUS and PP2C . D ( 2–9 ) pro:PP2C . D ( 2–9 ) -GUS seedlings treated with 10 μM IAA . Consistent with previous findings [42] , auxin induced PP2C . D7-GUS expression in the hypocotyls ( Fig 7A ) . Under our conditions , auxin induction of PP2C . D1pro:EGFP-GUS expression was not apparent in the hypocotyls . However , auxin strongly induced PP2C . D1pro:EGFP-GUS expression in the root elongation zone ( Fig 7B ) . We did not observe obvious auxin-induced expression of the PP2C . D2- , D3- , D4- , D5- , D6- , D8- , or D9-GUS reporters . To more precisely study the kinetics of auxin-induction of SAUR and PP2C . D gene expression , we examined the transcript levels of SAUR9 , 19 , 23 , and PP2C . D1-9 genes in 3-day-old light-grown wild-type seedlings treated with 5 μM NAA over a 2 h time-course by qRT-PCR . Auxin-induction of SAUR gene expression was observed within 10 minutes , and the induction peaked at 20–30 minutes ( Fig 7C ) . While auxin also induced PP2C . D1 and PP2C . D7 gene expression , the kinetics were noticeably delayed in comparison to the SAUR genes examined . PP2C . D1 was slightly up-regulated at 30 minutes and expression continued to increase throughout the 2 h time course , while PP2C . D7 expression was not elevated until 2 hours of auxin treatment ( Fig 7C ) . These results indicate that compared with the SAUR genes , PP2C . D1 and PP2C . D7 exhibit a delayed transcriptional response to auxin treatments . High temperature induces the expression of the SAUR19 family genes to promote hypocotyl growth [43] . We were curious whether high temperature might also affect PP2C . D gene expression . PP2C . D1pro:EGFP-GUS and PP2C . D ( 2–9 ) pro:PP2C . D ( 2–9 ) -GUS seedlings were shifted from 20 oC to 28 oC for 24 hours or 5 days and then stained alongside control seedlings maintained at 20 oC . Only PP2C . D2pro:PP2C . D2-GUS seedlings exhibited increased GUS staining in the hypocotyls ( Figs 7D and S9 ) . We therefore conducted a time-course to more carefully examine PP2C . D2 expression in PP2C . D2pro:PP2C . D2-GUS seedlings shifted from 20 oC to 28 oC . While no difference was observed at early time points ( 6 h ) , strong GUS staining in the hypocotyls and petioles of seedlings shifted to 28 oC was seen at the later time points ( Fig 7D ) . To gain additional evidence to support that high temperature upregulates PP2C . D2 protein levels , we examined PP2C . D2-GFP protein levels in PP2C . D2pro:PP2C . D2-GFP seedlings shifted from 20 oC to 28 oC . Increased PP2C . D2-GFP protein levels were detected in the shoots of seedlings shifted to 28 oC for 5 days ( Fig 7E ) . Together , these results indicate that high temperature upregulates PP2C . D2 protein levels . Since both the GUS and GFP reporters are translational fusions , this temperature-dependent increase could be the result of transcriptional or post-transcriptional regulation . To determine if this effect was the result of increased transcription , we examined PP2C . D2 transcript levels in light-grown wild-type seedlings shifted to 28 oC over a 120 h time-course by qRT-PCR . As previously reported [43] , high temperature induced the expression of SAUR19 and SAUR22 within 3 hours , and transcript levels remained elevated throughout the time-course ( Fig 7F ) . However , high temperature did not result in elevated PP2C . D2 transcript levels , suggesting that the observed temperature-dependent increases in PP2C . D2-GUS and PP2C . D2-GFP protein levels must occur post-transcriptionally . These results suggest that unlike the SAUR19 family genes , which display a rapid transcriptional response to high temperature , PP2C . D2 exhibits a delayed , post-transcriptional increase in protein abundance . Antagonistic regulation of cell expansion by SAUR proteins and the PP2C . D2 phosphatase led us to hypothesize that high temperature upregulation of PP2C . D2 protein abundance may represent an additional layer of control to prevent cell overexpansion conferred by increased SAUR gene expression , which may cause plant overgrowth . To test this hypothesis , we examined the hypocotyl growth responses of pp2c . d2-1 and pp2c . d2-2 loss-of-function mutants to high temperature . Two-day-old wild-type seedlings grown at 20 oC were shifted to 28 oC for 5 days , and hypocotyl growth was assessed . Indeed , both pp2c . d2 mutant lines exhibited an enhanced response to high temperature compared to wild-type controls ( Fig 7G ) . In contrast , the pp2c . d5 and pp2c . d6 single mutants exhibited temperature-dependent increases in hypocotyl length that were comparable to wild-type . Furthermore , while the pp2c . d2/5/6 triple mutant also exhibited enhanced elongation at high temperature , this increase was no more severe than that observed with the pp2c . d2 single mutants . These findings suggest that high temperature specifically upregulates PP2C . D2 protein abundance to prevent hypocotyl overgrowth .
Cell expansion is a fundamental cellular process that is essential for plant growth and development . Our prior work suggested that plasma membrane-localized SAUR proteins inhibit PP2C . D phosphatase activity to activate PM H+-ATPases , thereby promoting cell expansion [18] . Using the Arabidopsis hypocotyl system as a model , we demonstrate that the exclusively plasma membrane-localized PP2C . D2 , D5 , and D6 phosphatases negatively regulate hypocotyl growth ( Figs 2A and S5A ) . These proteins physically associate with PM H+-ATPases and negatively control cell expansion by dephosphorylating the penultimate threonine residue of PM H+-ATPases ( Fig 4 ) . Furthermore , we demonstrate that all three phosphatases can interact with SAUR19 in yeast 2-hybrid assays , that D5 co-immunoprecipitates with SAUR19 from plant extracts , and that the enzymatic activities of PP2C . D2 and D5 are inhibited by SAUR binding . While the pp2c . d2/5/6 triple mutant largely phenocopies GFP-SAUR19 overexpression plants , it is noteworthy that the hypocotyls of light-grown pp2c . d2/5/6 seedlings were still slightly shorter than those of GFP-SAUR19 seedlings ( Fig 3A and 3B ) . This suggests that in addition to the PP2C . D2 , D5 , and D6 phosphatases , additional PP2C . D family members may make minor contributions to the control of hypocotyl length . Consistent with this possibility , all family members except PP2C . D7 were strongly expressed in hypocotyls ( Fig 1A ) , and we found that all except PP2C . D8 could antagonize AHA2 function in the heterologous yeast expression system ( Fig 4D ) . Alternatively , it is also possible that SAUR19 has regulatory targets in addition to the PP2C . D phosphatases that may contribute to SAUR19-mediated cell expansion . Other than PP2C . D2 , D5 , and D6 , the remaining PP2C . D family members did not appear to influence cell expansion in hypocotyl growth ( S5A Fig ) . However , these proteins may regulate cell expansion to control the growth of other specific organs or developmental processes . Supporting this hypothesis , PP2C . D1 was found to regulate the differential growth at the apical hook in etiolated seedlings [18 , 24] . Our findings elucidate the contributions of individual PP2C . D phosphatases to cell expansion and provide insights into auxin-induced cell expansion via an acid growth mechanism . Prior computational analysis of PP2C . D proteins identified putative bipartite nuclear localization signals in all 9 family members and potential transmembrane spanning regions in PP2C . D1 , D3 , D4 , D6 , D7 , and D9 [41] . Our analysis of PP2C . D ( 1–9 ) pro:PP2C . D ( 1–9 ) -GFP reporters , as well as a prior study employing 35Spro:PP2C . D-GFP reporters [28] , correlate relatively poorly with these predictions . While PP2C . D2 , D5 , and D6 localized exclusively to the plasma membrane , the remaining PP2C . D family members localized to various cellular compartments , with D8 exhibiting mitochondrial localization , and D1 , D3 , and D4 exhibiting both nuclear and cytosolic localization ( Fig 2 ) . In the case of PP2C . D2 , D5 , and D6 , two of which lack a predicted transmembrane span , it is unclear how they associate with the plasma membrane . However , since all three physically interact with PM H+-ATPases ( Figs 4C and S7 ) , an attractive hypothesis is that they are recruited to the plasma membrane via their interaction with these H+ pumps . In regard to the non-plasma membrane family members , it is interesting to note that some SAUR proteins have also been shown to reside in the nucleus ( SAUR32 [44] and SAUR36 [45] ) and cytosol ( SAUR32 [44] , SAUR40 [46] , SAUR41 [47] , SAUR55 [45] , and SAUR71 [46] ) . However , the functions of SAUR proteins in these cellular compartments remain to be elucidated . We hypothesize that these SAUR proteins may regulate similarly localized PP2C . D proteins to control the phosphorylation status of their respective substrates . Identifying non-plasma membrane localized SAUR-PP2C . D regulatory modules and their substrates is an exciting area for future studies . Our loss- and gain-of-function studies demonstrate that the PP2C . D2 , D5 , and D6 phosphatases function as negative regulators to control diverse plant growth processes , including root , hypocotyl , leaf , flower , and silique growth . These proteins play a crucial role in elongation growth , such as hypocotyl and stamen filament growth ( Figs 3 and 5 ) . Stamen filament elongation growth during the late stages of stamen development is crucial for mature pollen to reach the stigma for a successful pollination . The yuc1/2/6 , tir1 afb1/2/3 , and arf6 arf8 mutants exhibit short stamen filaments [48–50] , indicating an essential function of auxin in stamen filament elongation growth . The auxin-induced SAUR63 family genes ( SAUR61-68 and SAUR75 ) , the likely downstream targets of auxin response factors ARF6 and ARF8 [50] , have been shown to positively regulate stamen filament elongation growth [34] , suggesting that these SAUR proteins may contribute to auxin-mediated stamen filament elongation growth . The PP2C . D2 , D5 , and D6 genes were all highly expressed in stamen filaments ( S1B Fig ) , and pp2c . d2/5/6 and PP2C . D5 overexpression flowers exhibited longer and shorter stamen filaments , respectively , than those of wild-type flowers ( Figs 3I and 5J ) . These results convincingly show that the PP2C . D2 , D5 , and D6 phosphatases negatively regulate stamen filament elongation growth and may therefore be important for the reproductive success of plants . It would be interesting to determine whether these three phosphatases physically interact with the SAUR63 family proteins to regulate stamen filament elongation growth . Our detailed phenotypic analyses of pp2c . d2/5/6 and GFP-SAUR19 plants indicate that pp2c . d2/5/6 plants phenocopy nearly all known phenotypes of GFP-SAUR19 overexpression plants , including increased cell expansion , hypocotyl and leaf growth , defective apical hook development and phototropic response , and elevated PM H+-ATPase phosphorylation and activity ( Fig 3 ) . These findings suggest that PP2C . D2 , D5 , and D6 phosphatases are the primary effectors of plasma membrane-localized SAUR proteins that regulate cell expansion . Differential growth is crucial for plant development and growth responses to environmental stimuli , such as light and gravity . Our studies implicate PP2C . D2 , D5 , and D6 as important regulators of differential growth , as the triple mutant exhibits defects in both phototropic bending and apical hook development . We previously reported that while the pp2c . d1 mutant exhibits defects in apical hook development , the pp2c . d2 , d5 , and d6 single mutants displayed apical hooks comparable to wild-type controls [18] . However , consistent with our hypothesis that these phosphatases function redundantly , etiolated pp2c . d2/5/6 seedlings exhibited clear defects in apical hook formation ( Fig 3J and 3K ) . Interestingly , while PP2C . D1 is specifically expressed on the inner side of apical hooks ( Figs 1C , 1D and 2B ) , PP2C . D2 , D5 , and D6 expression appears uniform across the hook ( Fig 1C ) . Together with the observed differences in subcellular localization ( Fig 2 ) , these findings suggest that PP2C . D1 and the three PM-localized phosphatases likely play distinct roles in modulating apical hook development . Furthermore , unlike PP2C . D1 , the uniform expression of the PM-localized family members across the apical hook suggests that regulatory proteins must be differentially expressed in order to achieve PP2C . D2/D5/D6-mediated differential growth . As auxin plays a prominent role in hook development , and auxin response is known to vary across the apical hook [51] , it seems likely that auxin-regulated SAUR proteins fulfill this function . Support for this contention can be found from prior gene expression studies of tropically-stimulated organs , which have revealed that multiple SAUR genes are differentially expressed in tropically-stimulated stems and hypocotyls , with expression being higher on the elongating side of the organ [37–39 , 52–55] . Since SAUR proteins inhibit PP2C . D activity , this differential pattern of SAUR expression would be expected to result in differential PP2C . D ( and consequently PM H+-ATPase ) activities across the organ , thus resulting in tropic bending . Indeed , differential apoplastic acidification was recently reported in gravistimulated Arabidopsis hypocotyls [20] . In the pp2c . d2/5/6 mutant , however , the primary effectors of SAUR function are missing , and consequently , defects in apical hook development and tropic bending result . While SAUR genes are rapidly upregulated in response to auxin , we did not observe auxin-mediated regulation of the majority of PP2C . D family members in either qRT-PCR or GUS reporter assays ( Fig 7C ) . PP2C . D1 and PP2C . D7 were notable exceptions , however , as both genes were auxin-inducible , albeit with delayed kinetics compared to SAUR genes ( Fig 7C ) . Although neither PP2C . D1 nor PP2C . D7 appear to play an important role in hypocotyl elongation under standard growth conditions , it is possible that delayed , auxin-induced expression could potentially function to attenuate SAUR-mediated growth regulation of other organs or processes , such as apical hook development . A more compelling case for increased PP2C . D expression functioning to attenuate SAUR-mediated growth was observed in our studies of high temperature-induced hypocotyl elongation . Arabidopsis seedlings grown under high temperature conditions exhibit elongated hypocotyls [56 , 57] . Our previous studies demonstrated that high temperature induces the expression of SAUR19 family genes to promote hypocotyl growth [43] . In this study , we found that high temperature also specifically upregulates PP2C . D2 protein levels ( Figs 7D , 7E and S9 ) . This increase appears to occur post-transcriptionally , as temperature did not affect PP2C . D2 mRNA levels . Compared with high temperature-induction of SAUR genes , the increase in PP2C . D2 protein abundance exhibited a delayed response to high temperature ( Fig 7D and 7F ) . We demonstrate that this increase in PP2C . D2 expression is biologically meaningful , as two independent pp2c . d2 mutant seedlings exhibited enhanced hypocotyl growth under high temperature conditions ( Fig 7G ) . These results provide strong support for the hypothesis that high temperature upregulation of PP2C . D2 protein abundance attenuates the hypocotyl growth conferred by SAUR proteins , preventing hypocotyl overgrowth . Our findings suggest a novel layer of regulation in high temperature-induced elongation growth . Determining the underlying mechanism ( s ) by which temperature affects PP2C . D2 protein abundance , such as enhanced translation or increased protein stability , is an exciting topic for future research that will further elucidate our understanding of high temperature-induced growth regulation .
Arabidopsis Genome Initiative locus identifiers for the genes employed in this study are as follows: SAUR19 ( At5g18010 ) , SAUR9 ( At4g36110 ) , SAUR22 ( At5g18050 ) , SAUR23 ( At5g18060 ) , PP2C . D1 ( At5g02760 ) , PP2C . D2 ( At3g17090 ) , PP2C . D3 ( At3g12620 ) , PP2C . D4 ( At3g55050 ) , PP2C . D5 ( At4g38520 ) , PP2C . D6 ( At3g51370 ) , PP2C . D7 ( At5g66080 ) , PP2C . D8 ( At4g33920 ) , PP2C . D9 ( At5g06750 ) , PP2C . I1 ( At2g25070 ) , PP2C . F9 ( At1g43900 ) , AHA2 ( 4g31090 ) , and AUX1 ( At2g38120 ) . Arabidopsis thaliana plants were grown under long-day conditions ( 16 h light/8 h dark ) under ~ 80 μEm-2s-1 fluorescent lighting at 20–22 oC unless stated otherwise in figure legends . All transgenic lines and mutants were in the Columbia ( Col ) ecotype . Seeds were surface sterilized in a solution containing 30% bleach and 0 . 04% triton X-100 and washed in sterile water . Seeds were cold treated for 2 to 3 days at 4 oC to synchronize germination . Seedlings were grown on ATS plates containing 1% sucrose and 0 . 5% Agargel ( Sigma-Aldrich ) . The ATS nutrient solution contained 5 mM KNO3 , 2 . 5 mM KPO4 , 2 mM MgSO4 , 2 mM Ca ( NO3 ) 2 , 50 μM Fe-EDTA , 70 μM H3BO3 , 14 μM MnCI2 , 0 . 5 μM CuSO4 , 1 μM ZnSO4 , 0 . 2 μM Na2MoO4 , 10 μM NaCI , and 0 . 01 μM CoCI2 . For medium acidification assays , 6–8 day-old light-grown seedlings were transferred from ATS medium to plates containing 0 . 04 mg/ml bromocresol purple ( BCP , Sigma ) and 0 . 5% Agargel with the pH adjusted to 6 . 5 with KOH , and incubated under long-day lighting as detailed above . Once color changes to the medium were visible , plates were imaged on a bed scanner . All statistical analyses were performed by analysis of variance ( ANOVA ) with the JMP Pro 13 . 1 software suite ( SAS Institute ) . Results of Tukey’s HSD ( honestly significant difference ) test were grouped by letters , with different letters indicating significant differences ( P < 0 . 05 ) . Sterilized seeds were cold treated for 3 days at 4°C in the dark , exposed to white light for 2 hours at 20 oC to induce seed germination , and then incubated for 4 days at 20 oC in the dark . Etiolated seedlings were photo-stimulated with unilateral blue light for various times , and the angles of hypocotyl bending were measured by ImageJ . Blue light ( 470 nm ) was provided by a SNAP LITE light system ( Quantum Devices ) . PP2C . D1-D9 genomic DNAs containing the promoters and coding sequences without the stop codons were amplified by PCR ( S1 Table ) and cloned into pENTR/D-TOPO using the pENTR directional TOPO cloning kit ( Invitrogen ) . In most cases , the entire intergenic region between the PP2C . D start codon and the previous gene was included . The length of upstream promoter sequence for each gene is listed in S1 Table . These PP2C . D inserts were recombined into pGWB203 ( GUS ) [58] , pGWB204 ( GFP ) [58] , and pEarleyGate 301 ( HA ) [59] using the Gateway LR clonase II enzyme mix ( Invitrogen ) to make PP2C . D ( 2–9 ) pro:PP2C . D ( 2–9 ) -GUS , PP2C . D ( 1–9 ) pro:PP2C . D ( 1–9 ) -GFP , and PP2C . D ( 2 , 5 , 6 ) pro:PP2C . D ( 2 , 5 , 6 ) -HA constructs , respectively . Similarly , 4426 bp of PP2C . D1 promoter sequence was cloned into pBGWFS7 ( EGFP-GUS ) [60] to make a PP2C . D1pro:EGFP-GUS construct . All binary vectors were introduced into Agrobacterium tumefaciens strain GV3101 ( helper plasmid pMP90 ) by electroporation . The floral dip method was used to transform Arabidopsis [61] . Transgenic plants were selected on ATS plates containing 0 . 01% herbicide Basta ( Bayer CropScience ) or hygromycin ( 25 μg/ml ) . GUS staining of plant tissues was performed at 37 oC in a solution containing 100 mM sodium phosphate ( pH 7 . 0 ) , 10 mM EDTA , 0 . 5 mM K4Fe[CN]6 , 0 . 5 mM K3Fe[CN]6 , 0 . 1% triton X-100 , and 1 mM X-Gluc ( 5-Bromo-4-chloro-3-indoxyl-beta-D-glucuronide cyclohexylammonium salt , Gold Biotechnology ) . After removing chlorophyll with 70% ethanol , GUS-stained tissues were cleared in a solution containing 20% lactic acid and 20% glycerol . GUS expression patterns were imaged with an Olympus SZX12 dissecting microscope using the SPOT Advanced imaging software . Plant microsomal proteins were prepared as previously described [62] . Co-IP and western blot assays were performed as previously described with the exception that the tris-buffered saline buffer ( TBST , 0 . 05% tween 20 , pH7 . 6 ) was used for western blots [18] . Far-western blot assays were performed as previously described [32] . GST-14-3-3 fusion proteins were detected by an HRP-conjugated anti-GST antibody ( GE Healthcare Life Sciences ) . Proteins were detected using the SuperSignal West Pico or West Femto Maximum Sensitivity Substrates ( Thermo Scientific ) . The lexA-based yeast two-hybrid system using the bait plasmid pBTM116 and the prey plasmid pACT2 [63] was used to examine SAUR19 and PP2C . D phosphatase interactions . pBTM116-SAUR19 and pACT2-PP2C . D plasmids were co-transformed into yeast strain L40ccU3 [MATa , his3-200 , trp1-901 , leu2-3 , 112ade2 LYS2:: ( lexAop ) 4-HIS3 , URA:: ( lexAop ) 8-lacZ , GAL4 , gal80] [63] , plated onto appropriate selection media , and grown at room temperature for 3 to 6 days . Saccharomyces cerevisiae strain RS-72 ( MATa , ade1-100 , his4-519 , leu2-3 , 312 , GAL1pro:PMA1 ) and the PMA1pro:AHA2 construct in the expression plasmid pMP1745 were previously described [40] . PP2C . D1-9 , PP2C . I1 , and PP2C . F9 full-length cDNAs were cloned into the NotI site of the expression vector pMP1612 [40] . All plasmids were introduced into yeast strain RS-72 by lithium acetate transformation . PMA1 complementation tests were performed as previously described [18] . The 6xHis-SAUR9 and 6xHis-PP2C . D1 constructs in the expression vector pET32 were previously described [18] . The full-length cDNAs of PP2C . D2 and PP2C . D5 in pENTR/D-TOPO were recombined into pET32-GW using the Gateway LR clonase II enzyme mix to make the 6xHis-PP2C . D2 and 6xHis-PP2C . D5 bacterial expression constructs . Expression and purification of His-tagged recombinant proteins and AHA2 dephosphorylation assays were carried out as previously described [18] . RNAs were prepared from seedlings using the RNeasy Plant Mini ( Qiagen ) or Nucleospin RNA Plant ( Macherey-Nagel ) kit , and an on-column DNase treatment was included to remove contaminating DNA . Two micrograms of RNA were used to synthesize cDNA using Moloney murine leukemia virus ( M-MLV ) reverse transcriptase ( Promega ) . qRT-PCR reactions were performed on the LightCycler System ( Roche Applied Sciences ) using the SYBR Green JumpStart Taq Ready Mix ( Sigma-Aldrich ) or StepOnePlus Real-Time PCR System ( Applied Biosystems ) using the Brilliant III Ultra-Fast SYBR Green QPCR Master Mix ( Agilent Genomics ) . Primers for qRT-PCR were previously described [18 , 36] , and results were based on three biological replicates . The full-length cDNA sequences lacking stop codons of PP2C . D2 , D5 , and D6 were cloned into pENTR/D-TOPO . These inserts were subsequently recombined into the pSPYNE and pSPYCE destination vectors [64] using Gateway LR Clonase II Enzyme Mix to generate BiFC expression constructs . AHA2 and AUX1 BiFC expression constructs were previously described [18] . All binary vectors were introduced into Agrobacterium strain GV3101 ( with the pMP90 helper plasmid ) by electroporation . BiFC assays were performed in an Nicotiana benthamiana transient expression system as previously described [65] . Leaves of ~ 5-week-old Nicotiana benthamiana plants were used for infiltration . The infiltration solution contained 10 mM MgCl2 , 10 mM MES-KOH ( pH 5 . 6 ) , and 150 μM acetosyringone . Fluorescent signals in leaf epidermal cells were observed three days after infiltration . Confocal microscopy was performed with a Nikon A1 spectral confocal microscope . | The plant hormone auxin is a major regulator of cell expansion , which is a fundamental cellular process essential for plant growth and development . The acid growth theory was proposed in the 1970s to explain auxin-induced cell expansion . However , the mechanistic basis of auxin-induced cell expansion via acid growth is poorly understood . Here , we investigated the functions of the D-clade PP2C ( PP2C . D ) family phosphatases in auxin-induced cell expansion as well as plant growth and development . The PP2C . D protein family is composed of nine members . Our findings demonstrate that the plasma membrane-localized PP2C . D2 , PP2C . D5 , and PP2C . D6 family members are the major regulators in auxin-induced cell expansion . These proteins physically associate with SAUR proteins and plasma membrane H+-ATPases to negatively regulate cell expansion . PP2C . D genes are broadly expressed and are crucial for a variety of plant growth and developmental processes , particularly elongation growth , such as hypocotyl and stamen filament growth . The results of our studies provide novel insights into auxin-induced cell expansion via an acid growth mechanism . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
]
| [
"plant",
"anatomy",
"plant",
"growth",
"and",
"development",
"enzymes",
"plant",
"embryo",
"anatomy",
"brassica",
"enzymology",
"phosphatases",
"hormones",
"membrane",
"proteins",
"developmental",
"biology",
"plant",
"science",
"model",
"organisms",
"plant",
"hormones",
"experimental",
"organism",
"systems",
"seedlings",
"plants",
"cellular",
"structures",
"and",
"organelles",
"flowering",
"plants",
"research",
"and",
"analysis",
"methods",
"arabidopsis",
"thaliana",
"plant",
"embryogenesis",
"plant",
"development",
"proteins",
"cell",
"membranes",
"biochemistry",
"plant",
"biochemistry",
"eukaryota",
"plant",
"and",
"algal",
"models",
"cell",
"biology",
"embryogenesis",
"biology",
"and",
"life",
"sciences",
"auxins",
"hypocotyl",
"organisms",
"fruit",
"and",
"seed",
"anatomy"
]
| 2018 | A subset of plasma membrane-localized PP2C.D phosphatases negatively regulate SAUR-mediated cell expansion in Arabidopsis |
Entamoeba histolytica is the etiological agent of human amoebic colitis and liver abscess , and causes a high level of morbidity and mortality worldwide , particularly in developing countries . There are a number of studies that have shown a crucial role for Ca2+ and its binding protein in amoebic biology . EhCaBP5 is one of the EF hand calcium-binding proteins of E . histolytica . We have determined the crystal structure of EhCaBP5 at 1 . 9 Å resolution in the Ca2+-bound state , which shows an unconventional mode of Ca2+ binding involving coordination to a closed yet canonical EF-hand motif . Structurally , EhCaBP5 is more similar to the essential light chain of myosin than to Calmodulin despite its somewhat greater sequence identity with Calmodulin . This structure-based analysis suggests that EhCaBP5 could be a light chain of myosin . Surface plasmon resonance studies confirmed this hypothesis , and in particular showed that EhCaBP5 interacts with the IQ motif of myosin 1B in calcium independent manner . It also appears from modelling of the EhCaBP5-IQ motif complex that EhCaBP5 undergoes a structural change in order to bind the IQ motif of myosin . This specific interaction was further confirmed by the observation that EhCaBP5 and myosin 1B are colocalized in E . histolytica during phagocytic cup formation . Immunoprecipitation of EhCaBP5 from total E . histolytica cellular extract also pulls out myosin 1B and this interaction was confirmed to be Ca2+ independent . Confocal imaging of E . histolytica showed that EhCaBP5 and myosin 1B are part of phagosomes . Overexpression of EhCaBP5 increases slight rate ( ∼20% ) of phagosome formation , while suppression reduces the rate drastically ( ∼55% ) . Taken together , these experiments indicate that EhCaBP5 is likely to be the light chain of myosin 1B . Interestingly , EhCaBP5 is not present in the phagosome after its formation suggesting EhCaBP5 may be playing a regulatory role .
Entamoeba histolytica is the etiological agent of amoebiasis ( intestinal as well as extra-intestinal ) , which results in a high level of morbidity and mortality worldwide , particularly in developing countries [1] , [2] . A number of studies have shown that Ca2+ and its binding proteins are centrally involved in amoebic pathogenesis and that cytolytic activity can be blocked by Ca2+ channel blockers or treatment with EGTA [3] . Genomic analysis of E . histolytica indicates the presence of 27 genes encoding multiple EF-hand calcium-binding proteins ( CaBPs ) [4] . The presence of such a large number of CaBPs suggests that this organism has a complex and extensive calcium signalling system [4] . One of the Ca2+ sensing proteins of E . histolytica , EhCaBP1 , has been extensively characterised , both structurally and functionally . EhCaBP1 was found to be involved in cytoskeleton dynamics and is associated with phagocytic cup formation in a Ca2+ independent manner [5] , [6] . The binding of Ca2+ to EhCaBP1 is necessary for the transition of phagocytic cups to phagosomes [7] . EhCaBP1 is recruited to phagocytic cups by the novel protein kinase EhC2PK [8] . The crystal structure of EhCaBP1 shows an unusual trimeric arrangement of EF-hand motifs [9] . The structure of the N-terminal lobe of EhCaBP1 displays a similar trimeric organization of EF-hand motifs as observed in the full length molecule . Lowering the pH to below physiological levels was shown to cause a trimer to monomer transition [10] . Moreover , various metal ions have been shown to impart flexibility and plasticity to the EF-hand motifs of EhCaBP1 [11] . We ( and others ) are systematically investigating the structure-function relationship of other calcium binding proteins of E . histolytica as well in order to understand their roles in amoebic biology and pathogenesis . Recently , an NMR structure of the calmodulin-like calcium-binding protein EhCaBP3 has been reported [12] . The N-terminal half of the molecule displays a structure similar to that of CaM , but no structure was derived for the C-terminal half of the molecule [12] . EhCaBP3 was found to be involved in the regulation of phagocytosis and cytoskeleton dynamics [13] . In addition to the studies of EhCaBP1 and EhCaBP3 , we have collected ( reported ) preliminary crystallographic data of EhCaBP2 [14] . Sequence analysis of the calcium binding protein 5 from E . histolytica ( EhCaBP5 ) indicates that its size ( 16 . 3 kDa ) and secondary structural arrangement are similar to those of CaM like proteins but it also suggests the presence of two calcium binding loops in two separate lobes . In CaM like proteins , two functional calcium binding EF-hand motifs usually exist side by side , and participate in calcium dependent target binding . The possible existence of two calcium binding sites in two separate lobes in EhCaBP5 prompted us to study the structure and function of this protein . We previously crystallized EhCaBP5 [15] , and here we report the structure determination and results from functional studies of this protein . We have determined the crystal structure in the calcium bound state at 1 . 9 Å resolution and have shown that EhCaBP5 interacts with the unconventional myosin ( myosin 1B ) , by surface plasmon resonance ( SPR ) . This interaction was further confirmed by a pull down assay and cellular co-localization using confocal microscopy . The results suggest that EhCaBP5 is involved in phagosome formation through interaction with myosin 1B . We conclude that phagocytosis in E . histolytica is regulated by a number of CaBPs including EhCaBP5 that regulates cytoskeletal dynamics and phagocytic cup formation with the help of myosin 1B , a process not observed in any other organism .
Molecular replacement with several calmodulin like proteins failed to give any solution , even though EhCaBP5 is , for example , 29% identical to potato calmodulin [16] . Instead , Se-Met labelled EhCaBP5 crystals were used to provide experimental phases and determine the structure ( see Methods for details ) . There is only one molecule present in the asymmetric unit . The final refined structure contains 131 residues , one calcium ion , 40 waters and 3 acetates; six residues from the N-terminus could not be modelled due to missing electron density . Overall , the molecule is divided in two globular lobes , where each lobe has four alpha helices connected by loops . The N-terminal lobe has one EF-hand motif with a calcium ion bound to the loop between two helices . The C-terminal lobe has two small anti-parallel beta strands along with four loops ( Fig . 1A ) . Signature residues of an additional EF-hand motif are found in the C-terminal lobe ( residues87 to 98 ) . In the crystal structure these residues form a loop between two helices similar to that of a typical EF-hand but density for Ca2+ is not observed . Consistent with the crystals structure , only one site in EhCaBP5 was found to bind Ca2+ by ITC [17] . Taken together , these results suggest that the C-terminal lobe of EhCaBP5 cannot bind Ca2+ . Nevertheless , the overall conformation of the Ca2+ bound N-terminal lobe is similar to that of the Ca2+ free C-terminal lobe , with an r . m . s . d . of 1 . 9 Å between these lobes . The coordination of Ca2+ observed in the EhCaBP5 crystal structure has only been seen in essential light chains ( ELC ) of myosin [18] , [19] and a calmodulin mutant [20] . In EhCaBP5 , the Ca2+ is coordinated by one carboxyl oxygen each of residues Asp16 , Asp18 , and Asp20 , the hydroxyl oxygen of Tyr22 , and two water molecules ( instead of one , as observed in CaM ) . The extra water also binds to the 12th position glutamate of the EF hand loop; this residue is usually critical in directly coordinating Ca2+ , but , in EhCaBP5 , it is too far away , at a distance of about 4 . 1 Å , and cannot directly coordinate instead it coordinates the Ca2+ via the intervening water molecule ( Fig . 1B ) . Overall , the Ca2+coordination geometry is octahedral in EhCaBP5 , instead of being pentagonal bipyramidal as is typically found in CaM . The octahedral geometry is more similar to that of Mg2+ coordinating in other calmodulin like proteins and essential light chain proteins . Among calcium binding proteins , EhCaBP5 displays the highest sequence similarity with potato CaM , with approximately 29% sequence identity . Moreover , EhCaBP5 is composed of two globular lobes similar to that of CaM . EhCaBP5 , however , differs from CaM regarding lobe composition . The central linker of EhCaBP5 is not a straight helix , but is broken in the middle , resulting in four separate helices in each lobe ( Fig . 1A ) . Both lobes interact with each other and appear as they are one over another . These differences may help explain the failure of the molecular replacement method for solving the EhCaBP5 structure as described above . There is also a striking difference between the ( Ca2+ bound ) EF-hand motifs of EhCaBP5 and that of CaM . The CaM EF-hand motif ( for example that in PDB code 1RFJ ) [16] adopts an open conformation after binding of Ca2+ , whereas the Ca2+ bound EF hand motif of EhCaBP5 is in a closed conformation . The r . m . s . d . between these EF-hand motifs is indeed relatively large , at 2 . 38 Å , and the interhelical angles of the EF-hand motifs of EhCaBP5 and CaM are 64 . 3 degrees and 89 . 3 degrees respectively , indicative of the closed and open conformations respectively ( Fig . 2A ) . The overall structure of EhCaBP5 is more similar to that of the essential light chain ( ELC ) of myosin than to CaM . Both EhCaBP5 and the ELCs have two 4-helix globular lobes connected by a small loop . Comparison with other Ca2+ binding proteins and ELC structures shows that the Ca2+ bound EF-hand motif of EhCaBP5 is particularly similar to that of squid myosin ELC [18] ( rmsd = 1 . 6 Å ) and to that of physarum myosin ELC [19] ( rmsd = 1 . 3 Å ) , adopting a closed conformation in each structure ( Fig . 2B ) . In these two ELC structures , as well as in EhCaBP5 , only one bound Ca2+ is seen in the N-terminal lobe . The closed EF-hand motif of squid ELC is due to the presence of an extra turn in the first helix and additional stabilizing interactions with the RLC [18] . In contrast , the EF-hand of the Physarum ELC structure lacks this extra helical turn , and in fact has a fully canonical loop structure , but is still observed in a closed state [19] . Such an unconventional mode of Ca2+ binding ( combination of closed EF-hand motif and canonical residues ) also occurs in EhCaBP5 . Also note that the conformation of the Ca2+ bound EF-hand motif of EhCaBP5 is very similar to that of the first EF hand motif of the trapped intermediate state of a CaM mutant ( rmsd = 1 . 08 Å; Fig . 2B ) ; in this mutant , the EF-hand motif is locked in a closed conformation by a disulphide bond , even though Ca2+ is bound [20] . Sequence alignment suggests that the residues of the first and second helices of the EhCaBP5 EF-hand motif are more hydrophobic than those in CaM and Physarum ELC . The extensive hydrophobic interaction between these two helices could be the reason for the observed closed state of the lobe and water mediated Ca2+coordination or intermediate state in EhCaBP5 . This indicates that Ca2+ binding energy is not sufficient to open the hydrophobic pocket . We tested whether EhCaBP5 can bind peptides representing IQ motif derived from myosin sequences as EhCaBP5 resembles ELC's , and since one of the two IQ motifs of myosin II binds ELC . E . histolytica genome encodes two myosins , myosin 1B and myosin II , containing one and two IQ motifs respectively . SPR was employed to carry out binding assays using the IQ motif peptide from myosin 1B and from ELC binding IQ motif of myosin II . The results indicate that EhCaBP5 does not bind the myosin II IQ motif while it does interact with the IQ motif from myosin 1B ( Kd = 2 . 4 nM ) ( Fig . 3A ) . To check the role of Ca2+ in this interaction , we performed this experiment in absence of Ca2+ . The result showed binding with Kd of 5 . 4 nM , suggesting that it takes place even in the absence of Ca2+ but with reduced affinity ( Fig . 3B ) . The binding of EhCaBP5 to myosin 1B IQ motif peptide is very specific , as neither BSA ( Fig . 3B ) nor EhCaBP3 ( Figure S1 ) was found to interact with immobilized IQ motif peptide . It is possible that EhCaBP3 interacts with myosin 1B through non IQ motif region . This binding seems to be stronger than CaM-Myo1c IQ motif interaction [21] . In order to further confirm the interaction between EhCaBP5 and myosin 1B , we carried out co-immuno precipitation using immobilized anti-EhCaBP5 antibody from the total cell lysate . The antibody precipitated myosin 1B along with EhCaBP5 even in the presence of EGTA confirming that Ca2+ is not required for EhCaBP5 to bind myosin 1B ( Fig . 4 ) . Immunofluorescence was used to investigate the localization of EhCaBP5 in proliferating amoebic cells . The results are shown in Fig . 5 . EhCaBP5 was found in the cytoplasm and no fluorescence signal was observed in the nucleus unlike EhCaBP3 . Since myosin 1B was shown to be involved in erythrophagocytosis [22] , further experiments were carried out to investigate whether EhCaBP5 is also involved in phagocytosis . We monitored EhCaBP5 localization during RBC uptake by E . histolytica to check the involvement of EhCaBP5 in phagocytosis . EhCaBP5 and myosin 1B were found in phagocytic cups based on analysis of fluorescence signals ( Fig . 6A , upper panel ) . Interestingly while myosin 1B was also found in the phagosomes ( denoted by asterisk ) as expected , EhCaBP5 was not seen , suggesting that EhCaBP5 is involved in the initiation phase of phagocytosis ( Fig . 6A , lower panel ) . Enrichment of actin was also observed in the phagocytic cups , as expected , and the superimposition of both EhCaBP5 and actin signals suggested that both proteins are co-localized in the phagocytic cups ( Fig . 6B ) . The results clearly show that EhCaBP5 , myosin 1B and actin are all colocalized in phagocytic cups . To check whether actin and EhCaBP5 also interact in vitro , we performed a co-sedimentation assay using F-actin and EhCaBP5 . No significant amount of EhCaBP5 was observed in the pellet fraction containing F-actin unlike EhCaBP3 , suggesting that participation of EhCaBP5 in phagocytosis follows a different path than that of EhCaBP3 . The results together clearly showed that EhCaBP5 is involved in amoebic phagocytosis by directly interacting with IQ motif of myosin 1B . The results shown so far suggest strongly that EhCaBP5 is involved in amoebic phagocytosis . In order to show if it is required , we determined erythrophagocytosis levels in cells where EhCaBP5 expression was down regulated by specific antisense RNA [23] . The vector used and details of different constructs are shown in Fig . 7A . The level of EhCaBP5 was significantly ( 62% ) reduced on tetracycline addition in the cells carrying antisense construct ( EhCaBP5AS ) as compared to the cells carrying only the vector ( Fig . 7B ) . This effect was specific as a control coactosin levels were monitored and the amount of coactosin did not change . EhCoactin is F-actin stabilizing protein , recently we have shown that overexpression of this protein results in arrest of phygocytic cup formation [24] . EhCaBP5 gene was over expressed using the cloned gene in the sense orientation ( EhCaBP5S ) the amount of EhCaBP5 increased by 25% in the presence of 10 µg/ml of tetracycline ( Fig . 7C ) . E . histolytica cells carrying all these constructs were then checked for erythrophagocytosis using a spectrophotometric assay . Cells expressing EhCaBP5 antisense RNA ( that is , in the presence of tetracycline ) displayed reduced ( 55% ) rate of phagocytosis as compared with cells carrying only the vector in the presence of tetracycline and the cells carrying EhCaBP5 antisense construct in the absence of tetracycline . On over expression of EhCaBP5 with the help of a sense construct , an increase of nearly 20% in erythrophagocytosis was observed as compared to cells without tetracycline or vector containing cells in the presence of tetracycline ( Fig . 7D ) . Phagocytic cup formation also showed similar pattern as that of phagocytic rates . The results are shown in Fig . 8 . Cup formation was reduced and delayed in cells expressing anti-sense RNA . A few cups were seen only after 10 min of incubation with RBC in antisense cells . In control cells generally cups were visible within a minute after addition of RBC . We also tested if EhCaBP5 is needed in the recruitment of myosin 1B . This was done by imaging myosin 1B in actively phagocytosing cells that are expressing anti-sense RNA of EhCaBP5 . There was no significant effect on myosin 1B staining in cells with reduced concentration of EhCaBP5 that is in presence of tetracycline in antisense construct carrying cells suggesting that EhCaBP5 is not needed in myosin 1B recruitment ( Fig . 9 ) . A theoretical model of the EhCaBP5-Myosin 1B IQ motif complex was generated using molecular docking simulations in order to predict conformational consequences of peptide binding as well as the details about the interaction between the protein and myosin 1B IQ motif peptide . The EhCaBP5 C-terminal domain adopts a more open conformation in the peptide bound model compared to native EhCaBP5 in absence of the peptide . It appears from our model that EhCaBP5 accommodates the IQ-motif peptide in the cleft , and N- and C-terminal lobes of EhCaBP5 move apart to wrap around the peptide . The model suggest that N-terminus of the peptide may interacts with the C-terminal lobe of EhCaBP5 , and the C-terminus of the peptide binds to the N-terminal domain of EhCaBP5 ( Fig . 10A ) . Probable interface residues of EhCaBP5 that may involve in interaction are F-15 , G-17 , E-27 , S-30 , R-33 , M-39 and D-117 with that of R-731 , G-729 , N-732 , K-725 , K-728 and R733 of the myosin peptide ( Fig . 10B ) . Superimposition of the native EhCaBP5 structure on the EhCaBP5-IQ motif model ( Fig . 10C ) indicates no global change upon peptide binding . The conformations of the N-terminal domains of the crystal structure and the Rosetta docked model are nearly identical , with an r . m . s . d . of 0 . 072 Å . The simulations , however , yielded about 15 degree reorientation of the C-terminal domain related to N-terminal domain , compared to the native structure , to accommodate the peptide resulting in a stretching out of the central loop connecting the two domains ( Fig . 10C ) . This has led to a change in overall length from 47 . 7 Å to 55 . 1 Å between native and peptide bound structures respectively . Moreover , the r . m . s . d . between the C-terminal domains of the native crystal structure and peptide bound model is 0 . 601 Å , reflecting greater predicted conformational changes within the C-terminal domain , as compared to within the N-terminal domain , upon binding of the peptide . These changes result in reorientations of the helices that help the molecule to have an open conformation needed to bind the peptide . The model described above was validated by performing SPR experiments using mutated/altered residues of IQ motif peptide ( sequence provided in material & methods section ) . Initially we mutated first two residues of peptide ( IQ to AA ) but we could not find any significant change in Kd value . Further we mutated Arg731 , Arg733 and Arg735 to Asp . The dissociation constant of mutated peptide with EhCaBP5 was calculated to be 4 . 1 mM as against 0 . 64 nM was observed with native peptide ( Fig . 11 ) . The observed dissociation constant shows that mutant IQ motif peptide binds to EhCaBP5 with less affinity and hence these positively charged residues Arg731 , Arg733 and Arg735 are possible key amino acids that are involved in interaction and affinity with EhCaBP5 , as indicated by the model .
One of the major interpretations of our structural studies is that the three-dimensional conformation of EhCaBP5 is more similar to that of myosin's ELC than to that of CaM , This was an unexpected finding as EhCaBP5 displays relatively high sequence similarity with CaM . This conformational similarity with ELCs is clearly seen on inspection as described in results and from the structure based alignment using the Dali server [25] . The most striking aspect of this similarity between EhCaBP5 and ELCs , but not with CaM , is that the EF-hand motifs of EhCaBP5 and ELCs are closed even in Ca2+ bound form . The EhCaBP5 EF-hand motif remains in a closed conformation after Ca2+ binding even though the Ca2+ coordinating residues are canonical . The closed EF-hand motif conformation of EhCaBP5 can bind to the heavy chain and can stabilize the closed state of the whole N-terminal lobe through cooperative interactions with or without calcium . In CaM , the decrease in energy resulting from the binding of Ca2+ compensates for the increase in energy accompanying the conformational change that opens up the hydrophobic pocket . However , there are more hydrophobic residues on the helices of EhCaBP5 than on the helices of CaM and there are two calcium binding loops in CaM compared to the one site in EhCaBP5; these features cause an increase in the energy that would be needed to open the hydrophobic cleft of EhCaBP5 . This is apparently not surmounted by binding of Ca2+ with the pocket remaining closed and the helices remaining stationary , the glutamate at the 12th position of the EF-hand motif is positioned too far to coordinate the Ca2+ . As a result , the Ca2+ bound EF-hand motif of EhCaBP5 is trapped in a so called intermediate state [20] . Our modelling of EhCaBP5 with myosin 1B IQ motif peptide suggests that EhCaBP5 adopts an extended conformation when it binds to myosin . Calmodulin like light chain of Mlc1p bound to IQ4 peptide of Myo2p also adopts an extended conformation , and it was expected that the extended conformation could mediate the formation of ternary complexes during protein localization and/or partner recruitment [26] . Therefore we expect that the extended conformation of EhCaBP5 with bound IQ motif may also allow for interactions with other molecular partners during various cellular processes . The observation that the structure of EhCaBP5 resembles that of ELCs , expands the opportunity for studying this myosin heavy chain binding class of proteins . The importance of ELCs in regulating the function of myosins is well known . However , most of these studies have been carried out in just a few systems , notably mammalian , drosophila and C . elegans . Recently , one of myosin II light chain ( CaBP20 or EAL50546 ) was identified [27] but there has been little information about ELCs and their regulatory role in myosin function in E . histolytica , especially on cellular myosin involved in phagocytic cup formation , and our functional studies of EhCaBP5 begin to address this issue . A number of evidences suggest that the binding partner of EhCaBP5 is myosin 1B but not myosin II . Assays using purified molecules as well as cell extract based assays and cellular colocalization have been used to demonstrate this interaction . Over-expression and suppression of EhCaBP5 also influences the rate of phagocytic cup formation . Since EhCaBP3 was also shown to bind myosin 1B , it is important to compare differential role of these two myosin 1B binding proteins . Our imaging experiments clearly showed that these two myosin 1B binding proteins have different specific functions , though overall both participate in phagocytosis . While EhCaBP3 stays on the phagosomes even after separation from membrane , EhCaBP5 is found till phagosomes are getting closed , but absent in phagosomes after separation . EhCaBP5 is not involved in recruitment of myosin 1B while EhCaBP3 is [13] and EhCaBP3 does not bind to IQ motif ( Figure S1 ) . Moreover , EhCaBP3 binds F-actin and myosin 1B in the presence of Ca2+ [13] , unlike EhCaBP5 that does not bind F-actin and does not require Ca2+ for interacting with myosin 1B . Furthermore , EhCaBP3 is also present inside the nucleus [13] , a feature not displayed by EhCaBP5 . These observations indicate both these calcium binding proteins are functionally different and that myosin 1B may be using different ELCs ( or binding proteins ) for carrying out different functions . Taken together , our structural studies show that EhCaBP5 resembles ELCs , and the functional studies indicate that it is likely to be the ELC of myosin 1B . E . histolytica has only two myosin in spite of high motility and tremendous high rate of phagocytosis . Other organisms , such as human has about 40 different myosin heavy chain genes , Dictyostelium discoideum , a closely related free living protist , encodes thirteen [28] . Therefore it is intriguing to understand how E . histolytica carries out all functions using only two myosins . We suggest that myosin 1B uses these two proteins as light chains to carry out different functions . If this is true then this helps partly to explain myosin paradox in E . histolytica . The mechanism by which it dissociates from myosin before phagosomes are closed is not clear . Particularly it is difficult to explain at present given the slow dissociation rate of bound IQ motif peptide , and the role of Ca2+ in this process . It is possible that other yet unknown regulatory proteins may be involved in this process . In this report we have attempted to delineate the function of the calcium binding protein EhCaBP5 using both structural and cellular approaches and showed that it is a myosin 1B binding protein and participates in phagocytosis . EhCaBP5 is one of a number of growing calcium binding proteins ( EhCaBP1 , EhCaBP3 , EhC2PK ) that have been recently identified to be involved in amoebic phagocytosis suggesting that Ca2+ has an important signalling role in phagocytosis .
Since the structure of EhCaBP5 ( accession number EAL46660 ) could not be solved by molecular replacement [15] , selenium-labelled protein was prepared to obtain phases . For preparation of selenium-labelled protein , E . coli BL21 ( DE3 ) cells containing EhCaBP5 plasmid were grown overnight in LB media . Cells were harvested by centrifugation at 4000 rpm for 10 minute . Harvested cells were washed with selenomethionine medium ( Molecular Dimensions ) twice , to take out residual LB medium , and then was suspended in the same media for inoculation for the further culture as described before [29] . For protein expression and purification we have followed same protocol as describe earlier for native protein [15] . Crystals of EhCaBP5 selenomethionine labelled protein were grown in similar conditions as were crystals of native EhCaBP5 [15] . SeMet- EhCaBP5 crystals were soaked in cryoprotectant solution consisting of 2 . 8 M sodium acetate , 0 . 1 M Bis-Tris pH 5 . 5 , and 20% glycerol . A single crystal was picked up in a cryo-loop and flash frozen in liquid nitrogen . A single wavelength anomalous dispersion ( SAD ) diffraction dataset was collected to 1 . 9 Å resolution at the Se edge ( λ = 0 . 9788 Å ) on a MARCCD 165 detector at the DBT-BM14 beamline of the European Synchrotron Radiation Facility ( ESRF , France ) . The peak dataset was then indexed , integrated , and scaled using the HKL2000 [30] . Data collection statistics are shown in Table 1 . The crystal belonged to space group C2 with unit cell parameters a = 70 . 55 Å , b = 44 . 45 Å , c = 47 . 73 Å , α = 90° , β = 108 . 9° , γ = 90° . Assuming one molecule of EhCaBP5 per asymmetric unit , the crystal volume per unit of protein mass 2 . 32 Å3/Da [31] , which corresponds to a solvent content of 47 . 3% . The sequence of EhCaBP5 consists of four SeMet residues and the positions of the Se atoms were determined using SHELXD program ( correlation coefficient , CC all/weak: 34 . 7/25 . 4; Patterson figure of merit , PATFOM 12 . 33 ) [32] . The initial phases were computed and partial model was built with SHELXE program as part of the HKL2MAP package [33] , [34] . This partial model was used as a starting point for iterative automated model building and rebuilding along with sequence docking using Auto Build program in Phenix software [35] , the remaining parts of the structure including side chains were modeled manually . The model was refined using the program REFMAC5 [36] and iterative manual rebuilding of the model was performed in COOT [37] . One Ca2+ atom was identified and included in the refinement . The translation-liberation-screw ( TLS ) displacement parameters were determined and TLS restrained refinement was performed [38] . For the final model , the Rwork is 18 . 7% and Rfree is 22 . 1% . The structure has good electron density ( Figure S2 ) and stereochemistry as indicated by program PROCHECK [39] with 96 . 1% of residues lying in the most favoured regions of the Ramachandran plot . The final refinement statistics are shown in Table 1 . The refined model of EhCaBP5 and structure factors was deposited in the Protein Data Bank under the accession code 4OCI . The E . histolytica genome codes for two myosin , myosin I ( accession number EAL48894 ) and myosin II ( accession number EAL51645 ) . Myosin I has one IQ motif and myosin II has two IQ motifs . To check the interaction between EhCaBP5 and myosin IQ motifs , we commercially synthesized ( obtained ) peptides of IQ motif of myosin 1B ( unconventional myosin ) ( IQKAWKGYRNRKR ) and second IQ motif of heavy chain myosin ( myosin II ) ( LQACARAFAARKHFS ) , which is expected to bind to ELC . For the binding study we used Biacore T200 apparatus ( Biacore , GE Healthcare ) at National Institute of Plant and Genome Research New Delhi , India . A total of 2000 resonance units ( RU ) of peptides were immobilized on a research grade S series CM4 sensor chip in 10 mM sodium acetate , pH 5 . 0 according to the manufacturer's amine coupling kit . After peptide immobilization , the surface was blocked with 1 M ethanolamine at pH 8 . 5 , followed by regeneration using 50 mM NaOH . The interaction experiments were performed using buffer containing 10 mM HEPES pH 7 . 4 , 150 mM NaCl and 0 . 2 mM Calcium chloride . We also performed interaction experiment in absence of Ca2+ and supplementing 5 mM of EGTA in the above buffer . Binding experiments were carried with different concentrations ( 125 , 250 , 500 , 750 , 1000 , and 2000 nM ) of EhCaBP5 in running buffer and injected at the rate 20 µL/min . For control we took bovine serum albumin ( BSA ) and EhCaBP3 at concentration 750 nM . The association kinetics for EhCaBP5 was monitored for 300 seconds and dissociation was monitored for the next 300 seconds . To validate EhCaBP5-IQ motif complex model , we obtained commercially synthesized two mutated Myosin 1B IQ motif peptide IQ-M1 A*A*KAWKGYRNRKR ( where IQ is mutated to AA ) and IQ-M2 IQKAWKGYD*ND*KD* ( Where R is mutated to D ) . The EhCaBP5 was immobilized on sensor chip up to 500 resonance units and native myosin 1B IQ motif peptide and mutated IQ motif peptide were passed as analyte at concentration of 25 , 50 , 75 , 100 and 125 mM . The data were recorded at 25°C and data analysis was performed using Biacore T2000 SPR Kinetics evaluation software . E . histolytica stain HM1: IMSS and the transformants were maintained and grown in TYI-S-33 medium as described before [40] . Hygromycin ( Sigma ) were added at 10 mg ml−1 for maintaining transgenic cell lines as indicated . Transfection was performed by electroporation . Mid-log phase cells were harvested and washed first by PBS and then cytomix buffer ( 10 mM K2HPO4/KH2PO4 ( pH 7 . 6 ) , 120 mM KCl , 0 . 15 mM CaCl2 , 25 mM HEPES ( pH 7 . 4 ) , 2 mM EGTA , 5 mM MgCl2 ) . The washed cells were then re-suspended in 0 . 8 ml of cytomix buffer containing 4 mM adenosine triphosphate , 10 mM glutathione and 200 µg of plasmid DNA . The suspension was then subjected to two consecutive pulses of 3 , 000 V cm −1 ( 1 . 2 kV ) at 25 µF ( Bio-Rad , electroporator ) . The transfectants were initially allowed to grow without any selection for 48 h . Selection was carried out by adding hygromycin B ( 10 µg ml−1 ) . pEhHYG-tetR-O-CAT shuttle vector was used for cloning of sense and anti-sense constructs . The CAT gene of pEhHYG-tetR-O-CAT [41] was excised using KpnI and BamHI and EhCaBP5 gene was inserted in its place in either the sense or the antisense orientation . The sequences of oligonucleotides used for making the above stated constructs are provided below , CaBP5_sense_FP-5′CGGGGTACCATGCAAAAACACAATGAAGAC-3′ CaBP5_sense_RP-5′GCGGGATCCTTACTTGAAAACAGTCATTAATTG-3′ CaBP5_anti sense_FP-5′CGCGGATCCATGCAAAAACACAATGAAGAC-3′ CaBP5_ anti sense _RP-5′CCGGGTACCTTACTTGAAAACAGTCATTAATTG-3′ Standard molecular techniques were used for making all these constructs . These clones were transfected as indicated above . Amoebic cells were labelled as described previously [5] . Cells grown at 37°C for 48 h were first washed with PBS and then with incomplete TYI-S-33 medium . The cells were then resuspended in the same medium and were allowed to grow on coverslips at 37°C for 10 min followed by fixation with 3 . 7% formaldehyde for 30 min , washed with warm 1× PBS and permeabilized with 0 . 1% Triton X-100 for 5 min . Additional treatment using chilled methanol ( −20°C ) for 3 min was carried out for staining myosin 1B . Permeabilized cells were then washed with PBS and quenched with 50 mM NH4Cl for 30 min at 37°C , followed by blocking with 1% BSA-PBS for 1 h . The cells were then stained with primary antibody for 1 h followed by Alexa Fluor 488 conjugated or TRITC conjugated anti-mouse secondary antibodies . F-actin was labelled with phalloidin using a similar protocol as above except the methanol step was omitted . Antibody dilutions used were: EhCaBP5 at 1∶200 , EhCaBP1 at 1∶200 , phalloidin ( Sigma; 1 mg/ml ) at 1∶250 , myosin 1B at 1∶150 , anti-rabbit or mice Alexa 488 ( Molecular Probes , Catalogue No . A-11008 or A-11001 ) at 1∶200 , anti-rabbit or mice Alexa 555 ( Molecular Probes , Cat . No . A-21428 or A-21422 ) at 1∶300 . The preparations were further washed with PBS and mounted on a glass slide using DABCO [1 , 4-diazbicyclo ( 2 , 2 , 2 ) octane ( Sigma ) 10 mg/ml in 80% glycerol] . The edges of the coverslips were sealed with nail-paint to avoid drying . Confocal images were visualized by using an Olympus Fluoview FV1000 laser scanning microscope . To quantify the red blood cells ( RBC ) ingested by amoebae , the colorimetric method of estimation was followed with little modifications [42] . Briefly , 1×107 RBCs were washed with PBS followed by TYI-S-33 and then incubated with 1×105 amoebae for different time points at 37°C in 0 . 5 ml culture medium . The amoebae and erythrocytes were pelleted and non-engulfed RBCs were lysed with cold distilled water and Centrifuged at 1000 g for 2 minutes . This step was repeated twice , followed by resuspension in 1 ml formic acid to burst amoebae containing engulfed RBCs . The optical density of the samples was determined by a spectrophotometry at 400 nm using formic acid as the blank . Immunoprecipitation was carried out as described previously [13] . Briefly , CNBr-activated Sepharose-4B beads ( Pharmacia ) were conjugated with anti-EhCaBP5 antibody . Crude immunoglobulins were collected from the immunized serum using 40% ammonium sulphate and subsequently dialysed in coupling buffer ( bicarbonate buffer ) . Usually , 10 mg immunoglobulin protein was added per gram of CNBr-activated Sepharose-4B beads . The resin was mixed gently for 18 h at 4°C . The conjugated Sepharose beads were incubated with E . histolytica lysate for 6 h at 4°C . The beads were then washed thrice with wash buffer ( 10 mM Tris-Cl ( pH 7 . 5 ) , 150 mM NaCl , 1 mM imidazole , 1 mM magnesium acetate , 2 mM β-ME and protease inhibitor cocktail ) . Ca2+ and EGTA were maintained throughout the process as required . After incubation the beads were washed sequentially with 60 mM Tris-Cl ( pH 6 . 8 ) , 100 mM NaCl and with 60 mM Tris-Cl ( pH 6 . 8 ) . The pellet was suspended in 2× SDS polyacrylamide gel electrophoresis ( PAGE ) buffer and boiled for 5 min followed by centrifugation for 5 min . The proteins were then analysed by western blotting . Analysis of the crystal structure showed that the conformation of the Ca2+ bound EF-hand motif of EhCaBP5 resembles that of myosin ELC , the coarse grained model of the EhCaBP5-peptide complex was obtained using the crystal structure of squid myosin , which contains its ELC and associated IQ motif-containing heavy chain ( PDB ID-3I5G ) [18] . The peptide bound conformation of EhCaBP5 was obtained by employing the flexible superimposition protocol of RAPIDO structural alignment software [43] . The course grained complex model of CaBP5-IQ motif used as the starting structure was then used for the molecular docking simulations with the IQ motif peptide using Rosetta FlexPepDock web server [44] . The Rosetta FlexPepDock protocol optimizes the protein-peptide complex using Monte-Carlo algorithm along with energy minimization [45] . In this study we used 200 models for refinement and chose the best model based on their Rosetta generic full-atom energy score ( Figure S3 ) . The images were prepared using Pymol [46] visualisation software . | Entamoeba histolytica is the etiologic agent of amoebiasis , a major cause of morbidity and mortality in developing countries . The genome of this organism encodes 27 EF-hand containing calcium binding proteins suggesting an intricate Ca2+ signalling system that plays crucial role in phagocytosis and pathogenesis . Calcium binding protein-5 ( EhCaBP5 ) is one of these CaBPs that displays sequence similarity with Calmodulin ( CaM ) but has only two possible calcium binding EF-hand loops in two separate domains . Interestingly crystal structure of EhCaPB5 showed more structural similarity with essential light chain ( ELC ) of myosin than that of CaM . The binding studies of EhCaBP5 with IQ motif peptides of myosins , showed that it interacts with IQ motif of unconventional Myosin IB . A number of experiments were carried out to show that EhCaBP5 indeed binds myosin IB and that this binding is Ca2+ independent . We also show here that EhCaBP5 participates in erythrophagocytosis and that its role in phagocytosis is different from that of EhCaBP3 , another myosin 1B interacting calcium binding protein of E . histolytica . Our results presented here and in a number of other reports point towards a unique phagocytic pathway involving a number of calcium binding proteins in E . histolytica . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"biochemistry",
"protein",
"interactions",
"molecular",
"complexes",
"proteins",
"protein",
"structure",
"biology",
"and",
"life",
"sciences",
"molecular",
"biology"
]
| 2014 | Crystal Structure of Calcium Binding Protein-5 from Entamoeba histolytica and Its Involvement in Initiation of Phagocytosis of Human Erythrocytes |
Oligomerization is a functional requirement for many proteins . The interfacial interactions and the overall packing geometry of the individual monomers are viewed as important determinants of the thermodynamic stability and allosteric regulation of oligomers . The present study focuses on the role of the interfacial interactions and overall contact topology in the dynamic features acquired in the oligomeric state . To this aim , the collective dynamics of enzymes belonging to the amino acid kinase family both in dimeric and hexameric forms are examined by means of an elastic network model , and the softest collective motions ( i . e . , lowest frequency or global modes of motions ) favored by the overall architecture are analyzed . Notably , the lowest-frequency modes accessible to the individual subunits in the absence of multimerization are conserved to a large extent in the oligomer , suggesting that the oligomer takes advantage of the intrinsic dynamics of the individual monomers . At the same time , oligomerization stiffens the interfacial regions of the monomers and confers new cooperative modes that exploit the rigid-body translational and rotational degrees of freedom of the intact monomers . The present study sheds light on the mechanism of cooperative inhibition of hexameric N-acetyl-L-glutamate kinase by arginine and on the allosteric regulation of UMP kinases . It also highlights the significance of the particular quaternary design in selectively determining the oligomer dynamics congruent with required ligand-binding and allosteric activities .
The biological function of proteins is usually enabled by their dynamics under native state conditions , which , in turn , is encoded by their 3-dimensional ( 3D ) structure . Unraveling this functional code has been the aim of many experimental and theoretical studies [1]–[9] . In particular the slow conformational dynamics of proteins in the micro-to-milliseconds time scale has been pointed out to be consistent with the changes in structure or domain/subunit movements observed between the substrate-bound and -unbound forms of enzymes [4]–[7] , [10] , and potentially limit the catalytic turnover rates of enzymes [11]–[14] . The quaternary structure of oligomeric proteins adds another layer of complexity to this code as the assembly of the subunits entails additional constraints while possibly inducing new types of collective motions . The structural hierarchy in oligomers indeed gives rise to a wide diversity of dynamical events [15] . For instance , in allosteric proteins , such as the paradigmatic hemoglobin [16] , [17] , the coupling between the internal dynamics of the subunits and the intrinsic ability of pairs of dimers to undergo concerted reorientations with respect to each other underlies the cooperative response to ligand binding [18]–[20] . Analysing the slow conformational dynamics thus emerges as a crucial step towards understanding the structure-function code in oligomeric proteins . Two classical models have been broadly used in the literature to interpret the conformational changes observed upon ligand binding: the Koshland-Némethy-Filmer ( KNF ) model [21] where the ligand ‘induces’ a conformational change in the allosteric protein , in line with the classical induced fit model , and the Monod-Wyman-Changeux ( MWC ) model [22] where the ligand selects from amongst those pre-existing conformers accessible by the intrinsic dynamics of the 3D structure . The former is usually a stepwise process , while the latter is all-or-none . The experimentally observed structural changes appear to result from a combination of intrinsic and induced effects: the intrinsic dynamics of the protein prior to substrate binding is essential to enabling cooperative changes in structure , while induced motions , usually more localized , help optimize and stabilize the bound conformers [4] , [23] . Protein-protein interfaces are usually characterized by their size , shape complementarity and hydrophobicity [24] , [25] . The dynamics at the interfacial residues are usually given little attention , although the functional significance of the structural changes triggered by complex formation or oligomerization is widely recognized . The interface between subunits often plays a key role in mediating the activity of each monomeric subunit [25] . Protein-protein interactions provide , not only thermodynamic stability to the folded state of the subunit in the complex ( or assembly ) , but also a new spectrum of collective motions . Furthermore , the oligomeric arrangement provides an efficient means of communication that may modulate allosteric regulation [19] . The present study focuses on the following questions: ( 1 ) Is the intrinsic dynamics of the component subunit modified by the oligomerization process , and if so , in which ways ? ( 2 ) What is the role of interfacial interactions and overall contact topology in the functional dynamics of the oligomer and , in particular , in signal transduction or allosteric communication ? The effect of multimerization on protein dynamics is investigated here in the context of the Amino Acid Kinase ( AAK ) family of enzymes . Members of this family have different degrees of oligomerization ( Figure 1 ) . Rubio and co-workers have significantly contributed to our current knowledge of this family of enzymes: they have resolved the X-ray structures of most family members [26]-[33] and suggested a shared mechanism of action on the basis of their sequence and folding similarities [28] . This mechanism was elucidated by our recent computational study of the softest modes of motion intrinsically accessible to different members of the AAK family of proteins [34] . The most exhaustively studied member of the AAK family is N-acetyl-L-glutamate kinase ( NAGK ) ( Figure 1A ) . NAGK phosphorylates the amino acid N-acetyl-L-glutamate ( NAG ) in the bacterial route of arginine biosynthesis . In many organisms , NAG phosphorylation is the controlling step of the route , as NAGK is feedback inhibited by the end product arginine . Rubio and co-workers [30] characterized the structures of two hexameric NAGKs ( from Thermotoga maritima ( Figure 1B ) and Pseudomonas aeruginosa ) that are cooperatively inhibited by arginine [35] . In Escherichia coli , NAGK ( EcNAGK ) is homodimeric and arginine-insensitive ( Figure 1A ) . Indeed , several studies have proven that the hexameric arrangement is a requirement for the cooperative inhibition by arginine [30] , [36] . The distinctive feature of this biosynthetic route in bacteria is that it produces N-acetylated intermediates , in contrast to mammals that yield non-acetylated intermediates . This turns NAGK into a potential target for antibacterial drugs by selective inhibition . Another member of the AAK family is carbamate kinase ( CK; Figure 1C ) . CK catalyses the formation of ATP from ADP and carbamoyl phosphate ( CP; a precursor of arginine and pyrimidine bases ) , and undergoes a substantial change in its structure upon substrate binding [37] . A third member is the hexameric UMP kinase ( UMPK ) ( Figure 1D ) . UMPK catalyzes the reaction ATP + UMP ADP + UDP to yield uridine diphosphate ( UDP ) . It is involved in the multistep synthesis of UTP , being regulated by the allosteric activator GTP and inhibited by UTP itself . Its monomer fold is very similar to the rest of family members , but presents a strikingly different assembly of the subunits that has not been explained so far . Notably , while the AAK family members do not exist in monomeric form , they share the same monomeric fold . This commonly shared monomeric fold is stabilized by oligomerization . The selection of a common monomeric fold in different oligomers suggests that that particular architecture possesses structure-encoded dynamic features that are exploited for enzymatic activity in oligomeric state . It is essential to analyze what the intrinsic dynamics of the monomeric units are , and to what extent , if any , they are maintained in the oligomeric state , or how they are coupled to , or complement , the dynamics of the biologically active ( oligomeric ) state . Calculations are thus performed for the monomeric fold alone as well as the monomer in the context of different oligomeric states , and the intact oligomers . As will be shown below , the oligomers do maintain some intrinsic dynamic features of the monomeric units , while the different assembly geometries of the monomers give rise to global motions uniquely defined for the particular oligomerization states . The method of analysis presented here is applicable to any protein that functions in different multimeric states . The effect of oligomerization on the dynamics of the component subunits can be experimentally examined provided that the protein exists in monomeric and different oligomeric states , which , in turn , may be controlled by environmental conditions [38] and few mutations at the protein surface [39] . However , such studies may be challenging in practice , and a computational examination emerges as an alternative promising tool . The most collective movements of biomolecular systems , also called the global modes of motions , can be determined using Elastic Network Models ( ENMs ) in conjunction with Normal Mode Analysis ( NMA ) at very low computational cost . A wealth of studies have shown the robustness of the global modes predicted by the ENMs ( e . g . , by the anisotropic network model , ANM [40] , [41] ) and their close relevance to experimentally observed structural transitions related to ligand binding [4]-[6] , [10] , [18] , [41]–[46] , or to the essential modes extracted from converged molecular dynamics ( MD ) simulations [47]–[49] . The global modes are the low-frequency modes extracted from NMA , also referred to as slow modes . They correspond to large-amplitude motions taking place at long timescales ( e . g . microseconds to milliseconds ) ; and they are also called soft modes due to their lower energy cost associated with a given level of fluctuation away from the equilibrium state , compared to other modes . Given their robustness and efficiency , ENMs are uniquely suited for exploring the collective motions and allostery in oligomers . Previous such studies have highlighted the significance of multimeric arrangement in defining the collective dynamics [50]–[54] . The present study adds new evidences to the role played by multimerization in defining functional dynamics . First , we contrast the low-frequency modes favoured by the EcNAGK and PfCK monomers to those preferentially selected by the corresponding dimers . Secondly , the modes of the monomeric and dimeric components of hexameric TmNAGK are compared to those collectively accessible in the hexameric form . Third , a detailed analysis of the softest modes accessible to the EcUMPK dimeric form is presented to shed light onto the role played by different dimeric assemblies found in the AAK family in selecting the functional motions of the family members . Overall , the different designs of interfaces and assembly geometries observed among the members of the AAK family are shown to practically define the collective modes that are being exploited by the oligomers for achieving their particular activities , including substrate binding and allosteric regulation .
How does the intrinsic dynamics of the monomeric subunits affect the oligomerization process or vice versa ? To what extent the intrinsic dynamics of the monomers prevail in the oligomers ? Or to what extent they are perturbed by oligomerization ? To analyse these issues , we have first compared the low-frequency ANM modes of the dimeric PfCK and EcNAGK with those of their respective monomers . The two enzymes exhibit close structural similarities ( Figure 2 ) . Their sequence identity is 24% , and their ATP-binding site and catalytic sites exhibit similar structural features . In fact , our previous comparative analysis of their collective dynamics showed that the slowest three ANM modes , which essentially modulate the opening/closure of the ATP-binding site , are commonly shared between these two enzymes; and they yield an overlap of 0 . 75 with the experimentally observed reconfiguration from open to closed state of NAGK [34] . The main structural difference between PfCK and EcNAGK , on the other hand , resides in their amino acid substrate binding site , and here we focus on the softest modes that control those sites . In EcNAGK , the β3–β4 hairpin serves as the lid of the NAG binding site and interlinks helices B and C , which are key components of the interface ( Figure 1A ) ; in PfCK ( Figure 1C ) , a subdomain protruding away from the interface serves as the lid of the CP binding site . This subdomain ( PS ) is formed by the strand β5 , helix αD and hairpin β6–β7 . Both lids exhibit significant conformational changes closely linked to substrate binding , as shown by the crystallographic studies performed by Rubio and co-workers [27] , [32] . Among the ANM modes that affect the substrate-binding sites , those simultaneously leading to closure/opening of the substrate-binding site in both subunits will be called symmetrical modes , and others , asymmetrical ( Figure 2 ) . In EcNAGK , the symmetrical opening/closure of the substrate-binding sites is enabled by the 5th mode ( red arrows in Figures 2B and 2D; see Video S1 ) , whereas the corresponding asymmetrical motion takes place in the 4th ( green arrows ) mode ( Video S2 ) . Note that our previous work [34] showed that ANM modes 1–3 were instrumental in accommodating the structural changes at the ATP-binding site , but had practically no effect on the NAG-binding site . This nicely illustrates how the enzyme takes advantage of different types of motions accessible to its native structure for achieving different types of functional motions . In mode 5 , the two β3–β4 hairpins ( Figure 1A ) , the lids of the NAG-binding sites , undergo an almost rigid-body rotation about the dyadic ( z- ) axis of the molecule while the ATP binding domains undergo smaller but coupled anticorrelated rotations . On the other hand , the asymmetrical motion ( mode 4 ) induces a translation along the y axis in both lids , along with the C-terminal part of the two helices B which are connected to the lids . No symmetric opening/closing of the lids is observed about the y-axis because these movements would be prohibited by steric clashes between the two B-helices ( blue arrows in Figure 2D ) . Rotational motions about the z-axis , on the other hand , are favored by the overall architecture of the dimeric enzyme . Indeed , tight interfacial interaction between the two B-helices is considered to be a key element for the stability of the dimer [28] . The interfacial region thus coincides with the central hinge site that mediates the opening/closing of the two monomers . This example emphasizes the effect of inter-subunit surface and topology on the character of the movements allowed/prohibited , or selected , in the oligomer . As to PfCK , the two substrate-binding subdomains are able to undergo both symmetric ( 1st and 3rd mode; see Video S3 ) and asymmetric ( 4th mode; see Video S4 ) motions because these two subdomains protrude away from the interface and their rotational rigid-body motions are not constrained by potential clashes between the adjacent B-helices . Indeed , the motion is parallel , rather than normal , to the plane defined by the two B-helices , and the two B-helices remain tightly packed and almost immobile in these modes . Notably , the global fluctuations of two PSs on PfCK dimer appear to modulate the access to the substrate-binding sites , suggesting a role in mediating substrate-binding . The selection of particular modes by EcNAGK for achieving its specific functions ( e . g . , modes 1 and 3 enabling ATP-binding; and mode 5 , substrate binding ) [34] raises the following question: is the rotation of the hairpins an acquired mode of motion originating from the topology of the dimer interface and not accessible to the monomer ? Or , is it an intrinsic dynamical ability of the monomer that is conserved and exploited in the dimer ? To address this issue , we compared the modes obtained for the isolated monomer with those of the monomer in the dimer , using the subsystem/environment coupling method described in the Methods . The monomer is the subsystem , and the second monomer stands for the environment in this case . For the sake of clarity , herein the modes that include the coupling to the environment are indicated with a superscript , i . e . , monomer ( dimer ) refers to the behaviour of the monomer within the dimer . The results are presented in Figure 3 ( and Supplementary Tables S1 and S2 ) . Therein the overlaps between the eight lowest-frequency modes accessible to the monomer in the isolated state ( y-axis ) and within the dimer ( x-axis ) are displayed for EcNAGK ( panel A ) and PfCK ( panel B ) , and Tables S1 and S2 lists the corresponding values . The orange-red entries along the diagonal in panel A demonstrate that the modes intrinsically accessible to the EcNAGK are closely maintained in the dimeric enzyme . Notably , both the order of the modes ( i . e . , their relative frequency and size , as defined by the respective eigenvalues ) , and their shapes are closely conserved . The picture is different in the case of the PfCK dimer ( panel B ) . While in EcNAGK all of the top-ranking seven modes are maintained with an overlap of 0 . 70 or above , in PfCK significantly fewer global modes favored by the isolated monomer are maintained , and with a weaker correlation and reordering of the modes . Thus , the PfCK monomer dynamics is strongly affected by dimerization . Examination of the individual modes showed that the monomer modes that induce high fluctuations at particular secondary structural elements such as the helix B and the β10–β11 hairpin ( shown in cyan in Figures 2A and C ) are practically absent in the dimer . As shown in Figure 2 these are key elements at the intersubunit interface , and dimerization imposes high constraints quenching their motion . The intersubunit surface of PfCK ( 2453 Å2 ) [27] is remarkably bigger than that of EcNAGK ( 1279 Å2 ) [28] . This higher surface area , and ensuing closer association of the two monomers , may be partly responsible for the larger perturbation of the intrinsic dynamics of the monomer upon dimerization in PfCK , compared to EcNAGK . Figure 2 and videos S3 and S4 in the Supporting Information demonstrate that the global motions preferentially undergone by the two PSs in the PfCK dimer induce conformational changes near the substrate-binding site; and Figure 3 shows that the global dimer dynamics departs from that of the isolated monomers . So , dimerization promotes in this case collective motions that affect substrate recognition and/or binding . The PS has been proposed to have evolved , together with the intersubunit interface , to play a key role in the specificity of CK for its substrate carbamate , as opposed to more abundant analogues , i . e . , acetate , bicarbonate or acetylphosphate [37] . This conjecture originally inferred from the examination of crystal structure alone is supported by our examination of PfCK dynamics . ANM global modes clearly indicate the ability of the PS to undergo movements toward the substrate-binding site , and the enhanced mobility at this particular region may indeed underlie the adaptability of CK to bind its substrate . The next case we studied is the hexameric form of the NAGK enzyme from Thermotoga maritima ( TmNAGK ) . The higher degree of multimerization of TmNAGK will permit us to contrast the dynamics of the whole enzyme with those of its dimeric and monomeric components . On the basis of the X-ray crystallographic structure , the hexameric arrangement of TmNAGK is considered to be a trimer of EcNAGK-like dimers [30] , herein called the AB dimer ( see Figures 1B and 4A ) . The dimeric scaffolds are interlaced by a mobile N-terminal helix , not present in the dimeric EcNAGK , and organized with a ring shape . An alternative dimeric building block being considered is the one constituted by the two monomers that interlink two adjacent AB dimers , herein called the AF dimer ( see Figure 4A ) . In the present study , we have compared the 20 lowest-frequency modes of the hexamer with those of the monomeric subunit and the two different dimeric building blocks . The results are presented in the panels B–F of Figure 4 . In each panel , the x-axis refers to the modes observed in the oligomer ( hexamer or dimer ) , and the y-axis refers to those intrinsically accessible to the components ( dimers or monomers ) that make these oligomers , e . g . , panel B compares the global modes of the AB dimer in the hexamer ( x-axis ) to those accessible to the AB dimer itself when examined in isolation ( y-axis ) . The comparative examination of these maps discloses two distinctive patterns: panels C and E reveal the conservation of global modes , in general , between the entities that are being compared , while panels B , D and F reveal that about ½ of the modes accessible to the substructures when examined in isolation are not represented in the assemblies . This behavior is clearly seen , and quantified , by the dashed lines on the maps , which represent a linear fit by weighted least squares regression to the entries that exhibit a correlation of 0 . 5 of higher . The dashed line in the former groups lies along the diagonal ( slope -1 . 04 and -1 . 01 in the respective panels C and E ) , whereas in the latter case , the slope varies as -1 . 81 ( panel B ) , -1 . 72 ( D ) and -1 . 44 ( F ) . Let us first examine the 1st group more closely: panel C essentially tells us that the monomers participating in the AB dimer maintain in the dimer their intrinsic dynamics favored by their monomeric architecture . As to panel E , it simply reflects that AF dimer in the hexamer behaves practically in the same way as in the isolated AF dimer , indicating that multimerization does not alter the global dynamics favored by the AF dimeric structure . In other words , the TmNAGK hexamer exploits the intrinsic dynamics of the AF dimer; and likewise , the AB dimer takes advantage of the structure-encoded dynamics of its monomers . Notably , the top four modes are conserved in this case with a correlation of more than 0 . 95 . This is in agreement with the high conservation of the monomer dynamics in the EcNAGK dimer , as pointed out in Figure 3A , given the structural and dynamical similarities [34] between the AB dimer and EcNAGK . We now turn our attention to the 2nd group . Here we see the dimer AB in the hexamer which is unable to sample several modes that are accessible to the same dimer in isolation ( panel B ) . Thus , the environment provided by the hexamer constrains the intrinsic dynamics of the AB dimer . Why is the AB dimer rigidified in the hexamer ? We note that in the hexamer , these EcNAGK-like ( AB ) dimers make close , interlacing interactions with the adjacent dimer by swapping their N-terminal helices and also making contacts with the C-domain , i . e . the interactions of AB-type dimers with the adjacent dimer through the AF interface impose topological constraints that impair several modes in the hexamer ( panel B ) . Likewise , the monomer in the hexameric environment is more restricted than the isolated monomer , such that many modes accessible to the isolated monomer cannot be effectuated in the hexamer ( panel D ) . Given the different degree of conservation of the dynamics of the AB and AF dimers within the hexamer ( panels B and E ) , we can add a complementary perspective to the structural view of TmNAGK as a trimer of EcNAGK-like dimers . The stronger conservation of the dynamics of the AF dimer supports a dynamical view of TmNAGK as a trimer of AF-like dimers . Finally , it is worth pointing out that the surface area of the AF interface ( 1186 Å2 ) is slightly smaller than that of the AB interface ( 1381 Å2 ) [30] . This might suggest that the monomeric modes would be more severely constrained in the AB dimer , but this does not hold true as explained above . The small difference in the surface area is therefore not sufficient to explain the observed behavior . The major determinant of accessible global motions is not the surface area but the topology of the interfacial contacts , or the overall shape/architecture of the dimer . In the present case , the overall architecture of the hexamer selectively hinders a number of global modes accessible to the AB dimer , while those of the AF dimer are mostly preserved . It is widely accepted that the size of the interface is closely linked to the thermodynamic stability of the oligomer [25] , [55] . The dynamics of the oligomer , on the other hand , is suggested by the present analysis to be predominantly controlled by the quaternary arrangement and contact topology of the subunits . The results discussed above focus on the preservation or the obstruction of the global motions of the subunits upon oligomerization . Nevertheless , in many cases , oligomeric proteins are subject to cooperative processes that regulate the biological activity . This raises the question whether such cooperative processes are linked to new modes of motion unique to oligomeric arrangement . TmNAGK is cooperatively inhibited by arginine in contrast to the dimeric EcNAGK and PfCK , which do not exhibit an allosteric regulation . The available X-ray crystallographic structure of TmNAGK represents the T state of the enzyme , which is bound to arginine . The apo form of the enzyme ( R state ) has not been structurally resolved , but the X-ray structure of the same enzyme from Pseudomonas aeruginosa ( PaNAGK ) serves as a suitable model for the R state on the basis of sequence and structural similarities [30] . Taking into account that the transition of TmNAGK between the R and T states is intimately linked to its allosteric regulation , those modes of motion that favor this conformational change will be the most functional . Therefore , the cumulative overlap of the lowest modes with the deformation vector between the R and T states has been calculated . Given that the T and R states correspond to proteins with different sequences , we have structurally aligned the two structures with DALI [56] and used the subsystem/environment coupling method ( see Methods ) to compute the ANM modes of TmNAGK , considering as subsystem those residues of TmNAGK structurally aligned to PaNAGK . Likewise , the deformation vector was calculated for the structurally aligned residues . Strikingly , a single non-degenerate mode ( 6th ) accessible to TmNAGK is found to describe 75% of the R↔T deformation ( see Figure 5D showing the cumulative overlap ) . A deeper analysis of this mode can shed light on the structural origin of the functionality of this enzyme . The aim is to ascertain whether this mode arises from the intrinsic dynamics of the subunits or is acquired in the hexameric state . Mode 6 is an expansion/contraction of the ring , accompanied by cooperative rotational and twisting motions of each monomer ( see Video S5 ) . The axis of rotation goes through each AF interface ( Figure 5A ) and performs an almost rigid rotation of the EcNAGK-like dimers ( Figure 5C ) . Residues close to these axes of rotation form minima in the mode fluctuations profile ( Figure 5B ) and belong to the AF interface . The axis involves a part of the N-terminal helix ( 6–20 ) of chains A and F , where the two helices interact tightly . Indeed , this interface stabilizes the hexameric arrangement and no NAGK dimer has been structurally characterized with an AF-like interface . The AF interface is unique to the hexameric arrangement . As shown in Figure 4 , the hexamer dynamics is affected by the intrinsic dynamics of the component subunits . Therefore , mode 6 could be associated with particular global modes accessible to the AB and/or AF dimers . We have examined the inter-residue distance variations maps induced by the low-frequency modes of the isolated AB and AF dimers to explore this possibility . AF dimer proves to be the major source of the rigid body movements of monomers observed in the hexamer ( see Videos S6 and S7 ) . The distance variation maps of the 1st and 4th modes of the AF dimer ( Figure S1 ) illustrate that the internal motions within a given subunit are negligible , but the relative movements between the two subunits are significant . The AF interface , thus , emerges as a key mechanical region that confers to the two linked subunits suitable flexibility to undergo functional changes in their relative orientations . This dynamic feature of the AF interface , whose size is smaller than the AB interface , is in accord with Hubbard and co-workers [57] , who stated that those interfaces that are not optimally packed may confer functional mobility to the oligomer . This inherent dynamical ability of the AF interface is therefore exploited in the hexameric arrangement to couple the rigid-body movements of the subunits , complementing their intrinsic internal dynamics . The topology of the AF interface appears to be evolutionary selected to provide two essential features for the functionality of the enzyme: ( 1 ) flexibility to allow for the cooperative reorientations of the dimers , which is inextricably linked to allostery , and ( 2 ) thermodynamic stability of the whole hexamer . Taking into account the crucial role of the AF interface and with the aim of providing further insights into the allosteric regulation of this enzyme , we considered the maximum likelihood pathway ( MLP ) for each combination of pairs of residues ( endpoints ) belonging to the respective chains A and F , and evaluated the fractional occurrence of each residue in the ensemble of MLPs ( see Methods ) . Figure 6A displays the percent occurrence of each residue , which also provides a measure of the relative allosteric potential of the residues . Peaks are observed at K17 , E18 , F19 , Y20 , K50 and Y51 ( ribbon diagram color-coded from blue ( peaks ) to red ( minima ) in Figure 6B ) . The significance of this first set in allosteric communication could be anticipated due to their location at the tightest part of the AF interface and proximity to the arginine inhibitor ( Figure 6B ) . However , our approach helps to identify other distal residues important for the communication , which behave as hubs . In particular , K196 and I162 channel most of the pathways to the AF interface via interactions with F19 ( and the arginine inhibitor ) and K50 , respectively . The communication across the AF interface can be summarized namely by two symmetric pathways distinguished by the MLP analysis: I162A → K50A→ Y51A→ K17F → E18F → F19F → K196F and its counterpart I162F →…→ K196A ( colored yellow and green in Figure 6B ) . Aromatic residues tend to be favored at protein interfaces [25] , and in this case , F19 and Y20 play a critical role . Not surprisingly , F19 is highly conserved among arginine-sensitive NAGKs [30] and , together with Y20 ( violet in Figure 6B ) , it establishes an efficient communication pathway of the form F19 ( A/F ) → Y20 ( A/F ) → Y20 ( F/A ) → F19 ( F/A ) . The structure of the monomeric subunit of EcNAGK is preserved among all family members , but the assembly geometry is less conserved . The arrangement of the monomeric subunits of NAGKs and CKs is strikingly similar , as shown above , but has significant differences with the assembly of UMP Kinases . Structurally , UMPKs are trimers of dimers in which the two helices that build the intersubunit surface of each dimer are parallel ( Figure 7C and D ) , whereas in NAGK ( and CK ) these helices at the interface make an angle of ∼65° ( Figure 7A and B ) . To our knowledge , a clear functional reason for this difference in monomer-monomer packing has not been reported so far . Although this difference has been argued to be necessary for hexameric assembly [58] , there might be another functional reason since TmNAGK is an example of a hexameric assembly that selectively adapts the EcNAGK-like dimer packing ( AB dimer ) . Here we compute the ANM modes of the UPMK dimer from Escherichia Coli ( EcUMPK ) in order to examine whether such a difference in packing geometry gives rise to significant changes in the global dynamics . The first mode of motion of the isolated EcUMPK dimer entails a rotational rigid-body movement with respect to an axis across the αC helices ( Figure 7 , panels C and D , and Video S8 ) . The anticorrelated motion of both subunits leads to an opening/closure movement of the whole dimer . This is in sharp contrast to the EcNAGK dimer dynamics , whose low-frequency modes do not exhibit rigid-body movements of the subunits . Does this dynamic feature of the EcUMPK dimer play a functional role ? Gilles and co-workers determined the X-ray crystal structure of EcUMPK complexed with GTP ( PDB code 2VRY ) [59] , which is an allosteric activator , and characterized a functional conformational change . They argued that GTP induces a rearrangement of the quaternary structure that involves a rigid-body rotation of 11° that opens the UMPK dimer . Strikingly , the first ANM mode predicted for the UDP-bound dimer describes the structural transition between the UDP- and GTP-bound forms . The overlap is outstandingly high ( 0 . 78 ) ( see Figure 8E for cumulative overlap ) . Moreover , it is worth pointing out that we have checked that this mode of motion is totally conserved in the hexamer ( see Figure S2 ) . Why does the different assembly in the UMPK dimer give rise to a normal mode with a rigid-body character not present in EcNAGK ? In UMPK the interface between the monomers is constituted mainly by two long parallel helices ( αC ) able to build a rotational axis that promotes an en bloc motion of both subunits . In contrast , the crossed orientation of the helices of NAGK ( ∼65° ) and the presence of other intersubunit contacts ( B-helices and β9–β10 hairpins ) hinders a rigid-body rotation of the two subunits . This suggests that the unique dimeric assembly of UMPK gives rise to a particular soft mode not present in other AAK family members . This example further indicates that the design of the interfacial contact topology and oligomerization geometry is crucial in defining the functional mechanisms of oligomers . In some cases , a single residue may significantly affect the contact topology at the interface and , thus , the allosteric regulation . This has been explored in the context of the UMPK analogue from Mycobacterium tuberculosis ( MtUMPK ) , for which crystallographic and site-directed mutagenesis studies have been recently conducted [60] . The X-ray structure of MtUMPK bound to GTP shows striking similarities to EcUMPK structure . Notably , this similarity is extended to their global motions: the lowest frequency ANM modes of the two structures exhibit an overlap of 0 . 97 . Given that the global modes of motion are fully determined by the overall shape of the protein , local perturbations are indeed unlikely to affect the low-frequency modes . Site-directed mutagenesis studies , on the other hand , show the importance of some residues in both the activity and the cooperativity of the enzyme . Among them , P139 was pointed out to to be a key residue in the allosteric regulation of the enzyme . P139 is located close to the trimeric interface where three GTP molecules are bound . What is the dynamical role of this residue ? The mean-square fluctuations profile obtained with the ANM shows that P139 occupies a position close to a local minimum ( a rigid part of the protein ) ( Figure 8A ) . Such regions usually play a key mechanical role for mediating collective changes in structure , and mutations at such positions may potentially affect the allosteric dynamics of the protein . We have analyzed the importance of P139 in mediating the allosteric communication among subunits A , D and F , which build one of the two trimeric interfaces where three GTP molecules are bound . We computed the communication pathways between GTP binding residues ( starting from subunit A and ending at subunits D and F ) and the percent contribution of each residue to MLPs , as done for TmNAGK . Figure 8 shows the trimeric interface color-coded according to the percent contribution in the same way as in Figure 6B . We note that the participation of P139 ( in yellow ) to these pathways is minimal ( note the red color in the backbone ) , but the adjacent residues Y137 and L138 are important mediators of inter-subunit communication via interactions with Q132 . This analysis suggests that the importance of P139 lies in constraining the orientation of nearby residues Y137 and L138 involved in inter-subunit signal propagation . The fact that this residue is highly restricted position in the global mode profile emphasizes its role in constraining the neighboring residues in a precise orientation pre-disposed to enable inter-subunit communication . The experimentally tested mutants ( P139A , P139W and P139H ) all showed a diminished allosteric regulation , but to different extents [60] . Further simulations at atomic scale might help explain the relative sizes of the effects induced by these mutations , but this is beyond the scope of the present work . It might be interesting to experimentally test the effect of mutations at L18 , Y137 and Q132 , since these residues emerge here as key elements enabling inter-subunit communication and they are distinctly restricted in the collective dynamics ( Figure 8A ) despite the relatively low packing density at the interface . To summarize , the present study reveals several dynamic features of oligomeric proteins by means of an ENM analysis of family members with different degrees of oligomerization . A common dynamic feature of the oligomers presented here is the conservation of the inherent dynamics of their monomeric or dimeric building blocks . The way these blocks are assembled in different oligomers confers different types of collective mechanisms unique to particular oligomerization geometries . Here are the main observations: In summary , the oligomers in the examined AAK family appear to selectively exploit the inherent dynamic abilities of its components , on the one hand , and favor coupled movements of intact subunits , on the other , to effectively sample cooperative movements ( soft modes ) that enable motions required for substrate binding and efficient allosteric responses . The architecture of the interfaces and the assembly geometry play an essential role in defining the most easily accessible ( or softest ) modes of motion , which in turn , are shown to be relevant to the functional mechanisms of the different oligomers , being presumably optimized by evolutionary pressure .
The low-frequency modes described by the NMA of different ENM variants [40] , [61]–[64] have proven to be robustly determined by the overall fold [7] , [65] , [66] and provide a consistent description of the conformational space most easily accessible to the protein [67] . Among them , we use here the most broadly used model , the anisotropic network model ( ANM ) [40] , [41] . In the ANM , the network nodes are located at the Cα-atoms' positions , and pairs of nodes within close proximity ( a cutoff distance of 15 Å , including bonded or non-bonded pairs of amino acids [41] ) are connected by springs of uniform force constant γ . The interaction potential of the molecule is given by ( 1 ) where M is the number of springs , and |Rij|-|Rij0| is the inter-residue distance with respect to the equilibrium ( crystal ) structure . The second derivatives of VANM with respect to residue displacements yield the 3Nx3N Hessian matrix H , the eigenvalue decomposition of which yields 3N-6 nonzero eigenvalues λk and eigenvectors uk corresponding to the frequencies ( squared ) and shapes of the normal modes of motion accessible to the examined structure . Numbering of modes in this work starts from the first mode with a nonzero eigenvalue . The cross-correlation between the displacements of residues i and j , contributed by mode k scales as ( 2 ) where the subscript ij designates the element of the matrix in square brackets . For i = j , equation ( 2 ) reduces to the square displacement of residue i in mode k . Clearly , lower-frequency modes ( smaller λk ) drive larger-amplitude motions . Conformations sampled upon moving along mode k are generated using ( 3 ) where R0 is the 3N-dimensional vector representing the initial coordinates of all residues and s is a parameter that rescales the amplitude of the deformation induced by mode k . The movies S1-S8 in the Supporting Information are generated using this equation with a series of different s values for selected modes of examined proteins . The degree of overlap between a conformational change Δr observed by X-ray crystallography and the structural change predicted by the ANM to take place along mode k is quantified by ( Δr · uk ) /|Δr| . Here Δr is the 3N-dimensional difference vector between the α-carbon coordinates of two different forms resolved for the same protein under different conditions ( e . g . , substrate-bound and -unbound forms of enzymes , or inward-facing or outward-facing forms of transporters ) . The cumulative overlap CO ( m ) between Δr and the directions spanned by a subset of m modes is calculated as ( 4 ) CO ( m ) sums up to unity for m = 3N-6 , as the eigenvectors form a complete orthonormal set of basis vectors in the 3N-6 dimensional space of internal conformational changes ( see Figures 5D and 7E ) The similarity between the conformational spaces described by two subsets of m and n modes , uk and vl , evaluated for two different systems can be quantified in terms of a double summation over squared overlaps as in Eq . 4 , among all mxn pairs of modes ( divided by m or n , depending on the reference set ) . The overlap O ( uk , vl , ) between the pairs of modes uk and vl calculated for different systems ( e . g . , Figure 3 ) is given by the inner product of the eigenvectors , i . e . , ( 5 ) Note that O ( uk , vl , ) is equal to the correlation cosine between the two N-dimensional vectors , since the eigenvectors are normalized . The change in a given inter-residue distance |R0ij| induced by a given mode k , , is given by the projection of the deformation induced by the kth mode onto the normalized distance vector , scaled by the inverse frequency , ( 6 ) Here ( uk ) i designates the ith super element ( a 3D vector ) of uk , and describes the relative displacement of the ith residue ( x- , y- , and z-components ) along the kth mode direction . Inter-residue communication has been suggested to play a key role in allosteric regulation and enzymatic catalysis [68] , [69] , and has been the subject of many computational studies [48] , [70]–[72] . Here we use a Markov model of network communication [73] , [74] to identify communication pathways . The interactions between residue pairs connected in the ANM are defined by the affinity matrix A , whose elements are aij = Nij/ ( Ni Nj ) ½ where Nij is the number of atom-atom contacts between residues i and j based on a cutoff distance of 4 Å , and Ni is the number of heavy atoms belonging to residue i . The density of contacts at each node i is given by . The Markov transition matrix M = {mij} , where mij = aij/dj , determines the conditional probability of transmitting a signal from residue j to residue i in one time step [73] . We define –log ( mij ) as the corresponding ‘distance’ . The maximum-likelihood paths ( MLPs ) for signal transfer between two end points are evaluated using the Dijkstra's algorithm [73] . In order to identify the residues that play a key role in establishing the communication between pairs of subunits , we considered the communication between all pairs of residues belonging to the two subunits of interest . In the application to the communication between the A and F subunits of TmNAGK ( Figure 6 ) , an ensemble of N2 = 2822 combinations of residue pairs ( endpoints ) have thus been considered ( each chain consists of N = 282 residues ) . For each pair , we evaluated the MLP and thus determined the series of residues taking part in the MLP . To quantify the contribution of a given residue to intersubunit communication , we counted the occurrence of each residue in the complete ensemble of MLPs . Figure 6 , panel A displays the resulting curve , peaks indicating the residues that make the largest contribution . In many applications the dynamics of a part of the protein ( subsystem , S ) may be of interest in the context of its environment ( E ) . The Hessian of the whole system is conveniently partitioned into four submatrices [75] , [76]: ( 7 ) where HSS is the Hessian submatrix for the subsystem , HEE is that of the environment and HSE ( or HES ) refers to the coupling between the subsystem and the environment . Inasmuch as the environment responds to the subsystem structural changes by minimizing the total energy , the effective Hessian for the subsystem coupled to the environment is ( 8 ) This approach has been advantageously employed in determining potential allosteric sites [77] and locating transition states of chemical reactions [78] . It will be used below in conjunction with the ANM for assessing the effect of oligomerization on the dynamics of monomeric and/or dimeric components ( subsystem ) . We examined four enzymes belonging to the AAK family ( Figure 1 ) : EcNAGK ( dimer ) , TmNAGK ( hexamer ) , PfCK ( dimer ) and EcUMPK ( hexamer ) . To this aim , we use the X-ray structures of EcNAGK in the open state ( PDB code: 2WXB ) , the arginine-bound TmNAGK ( PDB code: 2BTY ) , the ADP-bound PfCK ( PDB code: 1E19 ) and the UDP-bound EcUMPK ( PDB code: 2BND ) . All diagrams of molecular structures have been generated using VMD [79] . | Protein function requires a three-dimensional structure with specific dynamic features for catalytic and binding events , and , in many cases , the structure results from the assembly of more than one polypeptide chain ( also called monomer or subunit ) to form an oligomer or multimer . Proteins such as hemoglobin or chaperonin GroEL are oligomers formed by 2 and 14 subunits , respectively , whereas virus capsids are multimers composed of hundreds of monomers . In these cases , the architecture of the interface between the subunits and the overall assembly geometry are essential in determining the functional motions that these sophisticated structures are able to perform under physiological conditions . Here we present results from our computational study of the large-amplitude motions of dimeric and hexameric proteins that belong to the Amino Acid Kinase family . Our study reveals that the monomers in these oligomeric proteins are arranged in such a way that the oligomer inherits the intrinsic dynamic features of its components . The packing geometry additionally confers the ability to perform highly cooperative conformational changes that involve all monomers and enable the biological activity of the multimer . The study highlights the significance of the quaternary design in favoring the oligomer dynamics that enables ligand-binding and allosteric regulation functions . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
]
| [
"protein",
"interactions",
"enzymes",
"macromolecular",
"assemblies",
"globular",
"proteins",
"protein",
"structure",
"biophysics",
"simulations",
"biochemistry",
"simulations",
"proteins",
"enzyme",
"regulation",
"biology",
"proteomics",
"biophysics",
"macromolecular",
"complex",
"analysis",
"biochemistry",
"enzyme",
"structure",
"computer",
"science",
"computer",
"modeling",
"biophysic",
"al",
"simulations",
"computational",
"biology",
"macromolecular",
"structure",
"analysis"
]
| 2011 | Changes in Dynamics upon Oligomerization Regulate Substrate Binding and Allostery in Amino Acid Kinase Family Members |
Recessive skeletal dysplasia , characterized by joint- and/or hip bone-enlargement , was mapped within the critical region for a major quantitative trait locus ( QTL ) influencing carcass weight; previously named CW-3 in Japanese Black cattle . The risk allele was on the same chromosome as the Q allele that increases carcass weight . Phenotypic characterization revealed that the risk allele causes disproportional tall stature and bone size that increases carcass weight in heterozygous individuals but causes disproportionately narrow chest width in homozygotes . A non-synonymous variant of FGD3 was identified as a positional candidate quantitative trait nucleotide ( QTN ) and the corresponding mutant protein showed reduced activity as a guanine nucleotide exchange factor for Cdc42 . FGD3 is expressed in the growth plate cartilage of femurs from bovine and mouse . Thus , loss of FDG3 activity may lead to subsequent loss of Cdc42 function . This would be consistent with the columnar disorganization of proliferating chondrocytes in chondrocyte-specific inactivated Cdc42 mutant mice . This is the first report showing association of FGD3 with skeletal dysplasia .
Carcass weight , as a measure of meat yield , is an economically important trait in livestock . Due to its economic significance , several quantitative trait locus ( QTL ) mapping and genome-wide association studies ( GWAS ) have been conducted with the objective of identifying genes for improving meat production . These studies reveal that body length or stature is often times directly related to meat yield [1–3] . In humans , adult height is a highly polygenic trait affected by several hundreds of loci [4] , while some major loci have significant impact on body size in livestock [5 , 6] . In Japanese Black cattle , a previous GWAS detected three major loci for carcass weight [7] . Two loci , named CW-1 and CW-2 , correspond to PLAG1 [7] and NCAPG-LCORL [2] , respectively , both of which have been identified as loci influencing adult human height [8–10] and associated with body size-related traits in different cattle breeds and other livestock species ( reviewed in [11] ) . The third locus , named CW-3 , showed the largest allele substitution effect ( +35 . 0 kg ) among the three loci , while the Q allele frequency was the lowest ( 11 . 5% ) and detected in a specific line of Japanese Black cattle [7] . A QTL allele with a large effect size that is not near fixation , may be associated with an unfavorable trait . Examples include the Lys-232-Ala substitution in DGAT1 increases milk yield but reduces milk fat content in dairy cattle [12] , and a frame-shift mutation in MRC2 increases muscle mass in carriers but causes the recessive Crooked Tail Syndrome in Belgian Blue cattle [13] . In the case of CW-3 , skeletal dysplasia is observed more frequently following line breeding within the founding genetic line . The disease is characterized by joint- and/or hip bone-enlargement but the conditions are various . Here we show that CW-3 is linked inseparably with skeletal dysplasia and a non-synonymous variant of FGD3 causing a reduced activity of the encoding protein is a positional candidate QTN . The risk allele for skeletal dysplasia increased carcass weight in heterozygotes , most likely because of their taller stature and increased bone mass , while the homozygotes showed disproportionately narrow chest width causing an economic loss . The results explain the low CW-3 Q allele frequency despite its large effect on carcass weight .
The CW-3 QTL on bovine chromosome ( BTA ) 8 has been detected in seven Japanese Black paternal half-sib families of Sires I through VII ( Fig 1A ) , of which two ( Sires I and II ) were described previously [7] . An allele substitution effect ranged from +21 . 1 kg to +40 . 1 kg for carcass weight and the QTL explained 4 . 5% to 18 . 6% of the phenotypic variance within respective families ( S1 Table ) . The minimum overlapping region of the 95% confidence intervals ( CI ) for the QTL position was between MB065 ( 80 . 2 Mb , UMD3 . 1 ) and DIK2402 ( 95 . 0 Mb ) ( S1 Fig ) . Furthermore , the seven sires had a common ancestor , Sire X ( Fig 1A ) , and shared a ( hypothetical ) identical-by-descent ( IBD ) Q haplotype between IDVGA-52 ( 76 . 5 Mb ) and MS067 ( 90 . 2 Mb ) ( Fig 1B ) . The inheritance of the Q chromosome from Sire X to Sires I and II was evident ( Fig 1 ) . Interestingly , the q chromosome of Sire VII was also inherited from Sire X through Sire Y , and had a recombination event between MS089 ( 86 . 8 Mb ) and MS091 ( 87 . 5 Mb ) ; thus , the telomeric region distal to MS091 was identical to the Q chromosomes of Sires I and II ( Fig 1 ) . These results identify the telomeric end of the QTL interval at MS091 ( 87 . 5 Mb ) . On the other hand , the centromeric end of the QTL region was tentatively and conservatively determined at IDVGA-52 ( 76 . 5 Mb ) to include not only one ( Sire V , between MB065 and DIK1169 for body weight ) ( S1 Fig ) but also another 95% CI ( Sire II , between IDVGA-52 and DIK2402 ) [7] . The resultant critical region was an 11-Mb interval from IDVGA-52 ( 76 . 5 Mb ) to MS091 ( 87 . 5 Mb ) , which was covered by a shared Q haplotype ( Fig 1B ) . A GWAS using BovineSNP50 genotypes from 1156 Japanese Black steers also detected CW-3 [7] . The QTL was represented by a haplotype consisting of two single nucleotide polymorphisms ( SNPs ) but not by any single SNP [7] . To explore an SNP marker tagging the QTL , the BovineSNP50 genotypes were imputed to BovineHD genotypes using 651 steers as a reference population , followed by an examination for association . Twenty-two SNPs between 85 . 7 and 85 . 8 Mb were strongly associated with carcass weight ( p < 3 . 1 × 10–15 ) , of which the genotype of BovineHD0800025437 was experimentally validated to show 99 . 2% concordance ( 18 inconsistent alleles among 2312 alleles ) . Inclusion of BovineHD0800025437 genotype in the statistical model as a covariate resulted in the loss of all significant associations on BTA 8 , indicating that the SNP is in strong linkage disequilibrium with the causative variation for CW-3 ( Fig 2A ) . To further refine the CW-3 QTL region , a haplotype-based association analysis was performed using imputed BovineHD genotypes from the GWAS population . First , the 11-Mb CW-3 region was scanned for association with carcass weight by an approximately 1-Mb Q haplotype in a half-length sliding window . The most associated region , between 82 . 8 and 88 Mb , was then scanned by an approximately 500-kb-long window of the Q haplotype ( Fig 2B ) . The results narrowed the CW-3 region to a 3 . 3-Mb interval between 83 . 7 and 87 . 0 Mb ( Fig 2B ) . Skeletal dysplasia , characterized by joint- and/or hip bone-enlargement , has been known in a specific lineage of Japanese Black cattle ( Fig 3A and 3B ) . The disease causes economic damage to farmers because affected animals are bony and not fattened well . Since affected animals were produced from Sire II and its related sires , P and Q ( Fig 1A ) , 14 affected and 34 control animals from the three families were genotyped with the BovineSNP50 Genotyping BeadChip . Genotypes were used for homozygosity and autozygosity mapping using ASSHOM and ASSIST programs , respectively [14] . Both programs showed genome-wide significant signals on BTA8 ( p = < 10–4 ) ( Fig 4A ) , and the plot of p-values on BTA8 ( Fig 4B ) was similar to that of the GWAS for carcass weight [7] ( Fig 2 ) . The risk haplotype was on the same chromosome as the CW-3 Q haplotype in Sire II . Another 22 affected animals were collected from various sires and genotyped with BovineSNP50 . Of 36 affected animals , 29 ( 80 . 6% ) shared a homozygous 2 . 36-Mb risk haplotype between ARS-BFGL-NGS-108358 ( 84 . 18 Mb ) and BTA-52694-no-rs ( 86 . 54 Mb ) ( Fig 4C ) . Of the remaining seven affected animals , two did not possess the risk haplotype and five were heterozygous along BTA8 . The initial mapping contained three of the seven animals , while no significant regions were detected when the 7 affected and 34 control animals were subjected to ASSHOM ( p > 0 . 7 ) and ASSIST programs ( p > 0 . 11 ) . These seven animals may be erroneously diagnosed , because the disease phenotype is highly variable and the conditions are quantitative rather than qualitative as described below ( Fig 5B and S2 Table ) . The 2 . 36-Mb risk haplotype for skeletal dysplasia was encompassed by the 3 . 3-Mb Q haplotype of CW-3 , indicating they are closely linked ( Fig 1B ) . Although the IARS gene ( 85 . 3 Mb ) , whose missense mutation has been identified as the cause of hereditary perinatal weak calf syndrome in Japanese Black cattle [15] , is located within the 2 . 36-Mb region , the risk haplotypes were different for the two diseases . To search for candidate causative variations for skeletal dysplasia and/or CW-3 , three sires segregating the CW-3 QTL ( Sires II , V , and VII ) , three non-Q homozygous sires ( Sires D , I , and J ) , and a Q-homozygous steer ( Off-23 ) were subjected to targeted resequencing as described previously [7] . The 3 . 3-Mb CW-3 region contained 910 candidate QTNs ( 858 SNPs and 52 indels ) , including four non-synonymous and five synonymous SNPs . Coding regions that were not covered by targeted resequencing were read by Sanger sequencing , where four synonymous SNPs were detected ( S3 Table ) . Non-synonymous SNPs were located in FGD3 ( FYVE , RhoGEF , and PH domain containing 3 ) ( 85 . 8 Mb ) and PTPDC1 ( protein tyrosine phosphatase domain-containing protein 1 ) ( 86 . 8 Mb ) , and synonymous SNPs were located in BICD2 ( protein bicaudal D homolog 2 ) ( 85 . 7 Mb ) and FGD3 ( S4 Table ) . All non-synonymous SNPs showed strong association with carcass weight ( Fig 2C ) . The three non-synonymous SNPs in FGD3 showed complete linkage disequilibrium , while the linkage disequilibrium coefficient ( r2 ) between the non-synonymous SNP in PTPDC1 and those in FGD3 was 0 . 907 in the GWAS population . Using the non-synonymous SNP in PTPDC1 as a CW-3 marker , growth curves were compared between heterozygous ( n = 98 ) and homozygous q-steers ( n = 165 ) , indicating a highly significant effect of the Q allele on withers height ( p < 1 × 10–4 ) and body weight ( p < 0 . 003 ) during all test periods , while the effect on chest circumference was significant only at day 382 ( p = 0 . 039 ) and day 438 ( p = 0 . 020 ) ( Fig 5A ) . The results show that CW-3 primarily affects the stature as previously reported for other carcass weight QTL , PLAG1/CW-1 [3 , 16] and CW-2 [17] . BICD2 has been recently reported as the causative gene for spinal muscular atrophy in human [18–20] . The disease is characterized by lower-limb-predominant weakness [18–20] , which is different from skeletal dysplasia ( Fig 3A and 3B ) and CW-3 ( Fig 5A ) . Therefore we eliminated BICD2 from candidate genes and did not further examine synonymous SNPs in BICD2 . PTPDC1 was previously reported as one of 180 loci for adult human height [21] , while it constitutes the 237th locus with C9orf3 , PTCH1 and HABP4 in the most recent analysis that identified 697 variants clustered in 423 loci affecting adult human height [4] . The locus was defined as one or multiple jointly associated SNPs located within a 1-Mb window and in the 237th locus the majority of signals cluster in and around PTCH1 [4] that is known to relate to body size [22] . In bovine , there is an intra-chromosomal rearrangement between PTPDC1 ( 86 . 8 Mb ) and other three genes ( 82 . 6–84 . 6 Mb ) . To examine a possibility of PTPDC1 as a responsible gene for CW-3 , we first searched for the sires that have a recombination between FGD3 and PTPDC1 . Sire B [23] was found to be heterozygous for the non-synonymous SNP in PTPDC1 but had homozygous q alleles for the non-synonymous SNPs in FGD3 ( Fig 1B ) . An IBD analysis indicated that Sire B is heterozygous Q/non-Q between MS087 ( 86 . 0 Mb ) and MS091 ( 87 . 5 Mb ) of the CW-3 QTL interval ( Fig 1B ) , while a QTL for carcass weight was not detected in a previous QTL mapping study using 328 progeny of this sire [23] ( S1 Fig ) . These data strongly suggest that CW-3 maps centromeric to MS087 ( Fig 1B ) . Consistent with the results , Ptpdc1-deficient mice did not show any differences in body weight ( at 4–12 weeks of age ) or body length ( at 12 weeks of age ) ( S4 Table; 9–21 mice in respective genotypes ) , indicating that PTPDC1 is neither a height gene nor the cause for CW-3 . Although Ptpdc1 was highly expressed in testes ( S2 Fig ) , homozygous Ptpdc1-deficient mice were fertile ( five Ptpdc1-/- male mice when mated with C57BL/6J females , of which four produced offspring ) . The stronger association of the non-synonymous SNP in PTPDC1 than non-synonymous SNPs in FGD3 that are located centromeric to MS087 ( Fig 2C ) is probably due to sampling bias by chance . Only 19 animals had a haplotype with recombination between FGD3 and PTPDC1 , of which five had a q-Q and 14 had Q-q haplotypes . As for skeletal dysplasia , an affected animal has never been observed in the offspring from Sire B , suggesting that the causal mutation is located within the 1 . 8-Mb interval between 84 . 2 and 86 . 0 Mb ( Fig 1B ) . FGD3 located at 85 . 8 Mb is the only gene that has non-synonymous SNPs within the critical interval for skeletal dysplasia ( Figs 1B and 4C ) . The encoding protein functions as a guanine nucleotide exchange factor ( GEF ) for Cdc42 [24] . The five synonymous SNPs in FGD3 ( S3 Table ) were verified not to affect splicing ( S3 Fig ) , and they appeared to have no substantial effect on translation efficiency ( Codon usage database , http://www . kazusa . or . jp/codon/ ) . In contrast , the three non-synonymous SNPs in FDG3 caused loss of the start codon ( ATG to ACG ) and an amino acid substitution from His-171 ( CAC ) to Cys ( TGC ) . Since the Kozak consensus sequence is present at the second Met of the 17th amino acid residue ( S4 Fig ) , the FGD3 variant probably produces an N-terminal 16 amino acid-truncated protein . The His-171-Cys substitution is located in the GEF domain , where a His residue is conserved among mammalian species ( Fig 6A ) . In order to examine GEF activity of the wild and mutant FGD3 proteins , a pull-down assay was performed using a GST-PAK Cdc42/Rac interacting binding ( CRIB ) fusion protein . The wild and mutant FGD3 proteins were expressed as FGD3 ( SA ) in which two Ser residues of the DSGIDS motif were altered to Ala residues to protect them from proteasomal degradation [24] . As presented in Fig 6B and S4 Fig , the mutant protein showed reduced GEF activity . Both skeletal dysplasia ( Fig 3 ) and CW-3 ( Fig 5A ) are related to bone development . As a primary candidate gene , expression of FGD3 was examined in bone tissues . The results of the RT-PCR showed that bovine FGD3 was strongly expressed in the growth plate cartilage from a Holstein femur at 1 month of age but not expressed in ear cartilage or in bone marrow just under the growth plate ( Fig 6C ) . GAPDH was used as a positive control and COL2A1 , COL10A1 , and BGLAP were used as markers for chondrocytes , hypertrophic chondrocytes , and osteoblasts , respectively . FGD1 is the causative gene for human faciogenital dysplasia that is characterized by short stature , facial abnormalities , and skeletal and genital anomalies [25] . FGD1 was induced in micromass-cultured chondrocytes from ear cartilage , while FGD3 was not ( Fig 6C ) . Tibias were obtained from two FGD3 Cys-171 homozygous cows ( NLBC-1 and K in Table 1 ) and a non-carrier control cow ( T in Table 1 ) . Tibias from the homozygotes were longer and thicker than that of the non-carrier cow ( Fig 3C ) . In addition , medial malleolus and intercondylar eminence are more prominent in homozygotes . Overgrowth of the articular cartilage may be the cause of joint-enlargement , a phenotypic character of affected animals . To examine FGD3 gene expression precisely , in situ hybridization was carried out using a mouse femur . Fgd3 was expressed in osteoblasts and proliferating chondrocytes of the growth plate and articular cartilages ( Fig 6D ) , which were generally consistent with the regions where bone abnormalities were observed in the homozygous cows ( Fig 3C ) . To characterize the disease phenotypes , skeletal measurements were collected from risk allele-homozygous animals ( Table 1 and S2 Table ) . Their ages and conditions were varying and spinal or thoracic curvature was rare . To compare the measurements from the animals of different ages , deviations from normal growth curves were calculated . Six ( 25% ) of the 24 animals were recognized as having poor development . These animals were confirmed to be free from the IARS-risk haplotype or Val-79-Leu mutation of IARS [15] . Of the remaining 18 , 16 showed disproportionately narrow ratio of chest width to chest depth and 12 showed disproportionately wide ratio of thurl width to pin-bone width . Poor development and narrow chest width are correlated with economic loss to farmers . Although the phenotypes of skeletal dysplasia were thought to be congenital , some exceptions were found in this study . Four cows that were previously not recognized as affected at younger ages ( 14 . 5–22 months of age ) , showed a disproportion between chest width and chest depth at later ages ( 3 . 5–7 . 8 years of age ) ( Table 1 ) . In contrast , a proportion of thurl width to pin-bone width appeared unchanged irrespective of age ( Table 1 ) . To further examine the skeletal measurements of the risk-allele homozygotes , 332 offspring steers from the FGD3 His-171-Cys heterozygous sire , Sire R ( Fig 1A ) , were genotyped . Imputed SNP genotype showed that Sire R shared the CW-3 Q haplotype spanning the 11 Mb-interval from 76 . 5 to 87 . 5 Mb . The heterozygous calves were significantly taller ( p < 0 . 001 ) and heavier ( p < 0 . 01 ) than the non-carrier calves , but the chest circumferences were not different between these two groups ( p = 0 . 34 ) ( Fig 5B ) . The data were consistent with the results from the progeny tests ( Fig 5A ) , suggesting that the risk allele causes disproportional tall stature . The hypothesis was verified by the measurements of the risk allele-homozygous calves: they were taller ( p < 0 . 05 ) but had smaller chest circumference ( p < 0 . 01 ) than the non-carrier calves ( Fig 5B ) , indicating that disproportional tall stature is caused in an allele-dosage dependent manner . Body weight of the calves with homozygous risk alleles varied widely and was not significantly different from that of either the carrier or non-carrier calves ( Fig 5B ) . The smallest calf with homozygous risk alleles showed extremely small chest circumference that was classified as poor development , which may represent a severe case of the disease ( Table 1 ) . The carcass data at slaughter confirmed an increase in carcass weight as CW-3 in the carrier status ( p < 0 . 001 , Table 2 ) . The carcass yield estimates of heterozygous steers were lower than those of non-carrier steers ( p = 0 . 0013 ) , suggesting that the risk allele increases a ratio of bone to carcass weight . Carcass data of the risk allele homozygous animals could be traced for six of the nine animals . They were worse than those from heterozygous and non-carrier steers ( S5 Table ) . These results indicate that the risk allele causes disproportional tall stature in an additive manner . CW-3 is a result of an increase in height with unchanged width , while the homozygotes display skeletal dysplasia including poor development and disproportionately narrow chest width .
This study revealed that a QTL for carcass weight , CW-3 , is closely linked with recessive skeletal dysplasia . Skeletal measurements of calves revealed that the risk allele ( CW-3 Q allele ) causes disproportional tall stature in an allele dosage-dependent manner , suggesting that skeletal dysplasia and CW-3 are attributed to the same mutation . Only FGD3 has a non-synonymous coding variant within the critical region for skeletal dysplasia and the mutant protein showed reduced GEF activity , strongly suggesting that a mutation in FGD3 is causative . The mammalian FGD gene family consists of six members . Mutations in human FGD1 cause faciogenital dysplasia affecting multiple skeletal structures including short stature [26] . FGD6 was recently shown to regulate endosomal membrane recycling in osteoclasts [27] . FGD3 , similarly to other FGD proteins , functions as a GEF for Cdc42 [24] , however the specific role of FGD3 has been unknown . Lacroix et al . [28] reported that FGD3 is overexpressed in follicular thyroid tumors with PAX8-PPARγ1 rearrangement and localized at the lateral membrane but not at the basal or apical membranes . Because FGD3 is expressed in the growth plate cartilage ( Fig 6C and 6D ) , reduced GEF activity of FGD3 might reduce the active Cdc42 at the lateral membrane of chondrocytes , which may lead to columnar disorganization of chondrocytes seen in both limb bud mesenchyme-specific inactivated Cdc42 ( Cdc42fl/fl; Prx1-Cre ) mice [29] and chondrocyte-specific inactivated Cdc42 ( Cdc42fl/fl; Col2-Cre ) mice [30] . These cell type-specific inactivated Cdc42 mice indicate the essential role of Cdc42 in cartilage development during endochondral bone formation . However , the role of Cdc42 in postnatal limb skeletal growth and growth plate organization and function remains unclear because most of Cdc42fl/fl; Prx1-Cre mice die within a few days [29] and nearly all Cdc42fl/fl; Col2-Cre mice die within 1 day after birth [30] . Furthermore , another GEF for Cdc42 , such as Fgd1 , is also expressed in growth plate chondrocytes [26] , which may account for shorter limbs and bodies in Cdc42fl/fl; Prx1-Cre and Cdc42fl/fl; Col2-Cre neonate mice [29 , 30] . Therefore , generation and analyses of Fgd3-deficient mice may be crucial to verify the causality of Fgd3 for skeletal dysplasia . Disproportional tall stature is caused by activation of the C-type natriuretic peptide ( NPPC ) /NPR2 pathway . C-type natriuretic peptide binds to NPR2 in growth plate chondrocytes , which functions as a guanylyl cyclase to increase intracellular cGMP level [31] . The increase in cGMP level further activates cGMP-dependent protein kinase II and seems to promote the accumulation of extracellular matrix in the growth plate [32] . A spontaneous loss-of-function mutation of mouse Npr2 causes severe disproportionate dwarfism [33] , while a gain-of-function mutation of NPR2 producing excessive cGMP causes human overgrowth disorder [34] . The transgenic mouse model expressing the gain-of-function mutant Npr2 in chondrocytes exhibits bone deformities , which , depending on the expression levels of the transgene , include elongation of the spine with severe kyphosis or elongated spinal and tail vertebrae and phalanges with mild kyphosis [34] . NPR3 functions as a natriuretic peptide clearance receptor and the NPR3 inactivated mice also show a disproportionate tall stature [35] . The diverse phenotypes of the mice expressing the gain-of-function mutant Npr2 resemble various conditions of bovine skeletal dysplasia . Thus , a reduced activity of bovine FGD3 may induce growth plate disorganization , which may in turn lead to activation of the NPPC/NPR2 pathway . NPR2 is also located on BTA8 but its genomic position ( 60 . 4 Mb ) is apart from the critical region for skeletal dysplasia ( Fig 4C ) . Fasquelle et al . [13] showed that a frame-shift mutation in MRC2 causing the recessive Crooked Tail Syndrome gives desired characteristics such as increased muscularity in the carrier status in Belgian Blue cattle . Likewise , CW-3 Q allele increases carcass weight in the heterozygous state . Because CW-3 heterozygous bulls tend to be preferentially selected due to enhanced longitudinal growth , marker-assisted breeding will be useful to avoid an increased frequency of skeletal dysplasia .
This research was approved by the National Livestock Breeding Center for Animal Research ( H26-5 ) and conducted in accordance with the Institutional Animal Care and Use Committee Guidelines from National Livestock Breeding Center . QTL analyses were performed with the interval mapping method using a linear regression model for half-sib families , as described previously [7] . Slaughter year and age ( day ) were included as fixed effects in a model . Marker locations were obtained from the Shirakawa-USDA linkage map [36] . Genomic sequences were examined to identify microsatellites . Primers targeting microsatellites or SNPs were designed using Primer 3 ( http://bioinfo . ut . ee/primer3/ ) [37] . The UMD3 . 1 assembly was used for genomic positions . Genotyping of microsatellites was performed using polymerase chain reaction ( PCR ) with a fluorescently labeled reverse primer , followed by electrophoresis using an ABI 3730 DNA analyzer ( Applied Biosystems , Foster City , CA , USA ) and analysis using GeneMapper software ( Applied Biosystems ) . The sires and their offspring were genotyped to determine the phase of the sires’ chromosomes . For genotyping of SNPs , direct sequencing of the PCR products was performed using BigDye Terminator v . 3 . 1 Cycle Sequencing Kit ( Applied Biosystems ) , followed by electrophoresis using an ABI 3730 DNA analyzer . The primer sequences are shown in S6 Table . The GWAS population consisted of 1156 Japanese Black steers whose carcass weight had higher ratios of both extremes than the ratio observed in collected samples ( > 27 , 500 ) but was normally distributed [7] . The BovineSNP50 ( Illumina , San Diego , CA , USA ) genotypes on BTA8 of the GWAS population were imputed to BovineHD ( Illumina ) genotypes using Beagle 3 . 3 . 2 . [38] . The BovineHD genotypes of 651 Japanese Black steers that passed our quality-filter [7] were used as a reference population . Association of imputed genotypes with carcass weight was examined using EMMAX software [39] . The IBS matrix was made using BovineSNP50 genotypes as previously described [7] . To examine imputation accuracy , an SNP , BovineHD0800025437 , was genotyped for the 1156 steers by direct sequencing of the PCR products ( S6 Table ) . The sires segregating CW-3 ( Sires I-VII ) , their common ancestor ( Sire X ) , and an offspring from Sire VI ( Off-6 ) were genotyped with BovineHD Beadchips ( Illumina ) , followed by imputation as described above . Since imputed genotypes were obtained as phased haplotypes , the CW-3 Q haplotype consisting of BovineHD SNPs was obtained as a shared haplotype among the sires and confirmed by the Off-6 genotype that possessed homozygous Q alleles between 77 . 8 and 100 . 9 Mb . For a haplotype-based association analysis , one out of every nine SNPs was taken from the phased BovineHD genotypes of the GWAS population to reduce the number of SNPs that constitute haplotypes . The average interval of the SNPs was 50 kb . Haplotypes consisting of 20 contiguous SNPs , whose length was estimated to approximately 1 Mb , were divided into Q or non-Q haplotypes . Association of Q haplotypes with carcass weight was examined with a sliding window of 10 SNPs using EMMAX software [39] as described above . For an analysis of approximately 500-kb haplotypes , the first and fifth of every nine SNPs were used . The initial mapping was done with 14 affected and 34 control animals from three families: 5 affected and 10 control offspring from Sire II , 5 affected and 10 control offspring from Sire P , and 4 affected and 12 control offspring from Sire Q . Sires P and Q were also used as control animals . Both affected and control animals included males and females . They were genotyped with BovineSNP50 Beadchips and submitted to homozygosity and autozygosity mapping using ASSHOM and ASSIST programs , respectively [14] . Further , 22 affected animals were collected from various groups of sires and genotyped with BovineSNP50 Beadchips . A total of 36 affected animals were used to search for a shared homozygous region on BTA8 . Three sires segregating the QTL ( Sires II , V , and VII ) , three non-Q-homozygous sires ( Sires D , I , and J ) [7] , and a Q-homozygous offspring from Sire VI ( Off-23 ) were subjected to sequence capture ( NimbleGen custom array ) , followed by resequencing ( Illumina GAIIx , 40-bp paired-end run ) . This experiment was performed in parallel with the targeted resequencing of the CW-1 region [7] . Non-Q-homozygous sires were defined as those that harbored homozygous q alleles at BovineHD0800025437 and did not segregate a carcass weight QTL on BTA 8 in the half-sib analyses using more than 236 offspring per family . Targeted region included an 11-Mb interval of the CW-3 region from 76 . 5 to 87 . 5 Mb on UMD3 . 0 . Obtained putative sequence variations were filtered by: ( 1 ) heterozygous in the three sires segregating CW-3 QTL , ( 2 ) homozygous in the three non-Q- and Q-homozygous animals , and ( 3 ) opposite alleles between the Q- and non-Q-homozygous animals . Coverage of the coding regions in the 3 . 3 Mb-interval between 83 . 7 and 87 . 0 Mb was checked in each animal . The regions with fewer than four reads in either of the animals were subjected to Sanger sequencing ( S3 Table ) . Two hundred and sixty-four steers from 35 sires , which were used for progeny tests at Shimane Prefectural Livestock Technology Center from 2002 to 2011 , were genotyped for non-synonymous SNP in PTPDC1 ( S3 Table; bPTPDC1_e6_FV in S6 Table ) . There was only one steer that had homozygous Q alleles ( = G ) and was excluded from the analysis . The measurements of withers height , body weight , and chest circumference were interpolated by cubic spline at 4-week-intervals starting at 270 days of age . Significance of the difference between heterozygous and homozygous q-steers was tested using Student’s t-test . Skeletal measurements of calves , withers height , body weight , and chest and abdominal circumferences , were collected at the stock market in Miyagi Prefecture from October in 2008 to March in 2009 . Hair roots were collected for DNA samples . Of those , 333 offspring steers from Sire R were genotyped for the SNPs encoding His-171-Cys in FGD3 ( S3 Table; FGD3_e2_HC in S6 Table ) . Sire R was heterozygous for the SNPs . The age of the calves was 295 ± 21 days . The phenotypic values were compared between genotypes using ANOVA . Approximately half of the genotyped animals could be traced for the carcass data at time of slaughter . Because the carcass data of the FGD3 Cys-171 homozygotes were significantly worse than those of other genotypes , they were excluded from the subsequent linear regression analysis . Association of the risk allele with carcass traits was examined using a linear regression model including slaughter age as a covariate . Sire R was genotyped using Axiom Genome-Wide BOS1 Array Plate ( Affymetrix , Santa Clara , CA , USA ) , followed by imputation to Bovine HD genotype as described above . Heterozygous Ptpdc1-deficient mice ( TF0596 ) were obtained from Taconic Knockout Mouse Repository ( Taconic Biosciences , Inc . , Hudson , NY , USA ) . The targeted allele was deleted from exon 4 to exon 7 of the Ptpdc1 gene and replaced by LacZ and Neo genes . The heterozygous males were backcrossed into C57BL/6J females ( CLEA Japan , Tokyo , Japan ) four times and the resultant heterozygous males and females were crossed to obtain littermates with respective genotypes . Genotyping was performed by PCR analysis . The PCR reaction was done using TAKARA LA Taq and GC buffer I ( Takara , Tokyo , Japan ) in a single reaction tube containing the following three primers: Neo_new , 5’-TCGCCTTCTTGACGAGTTCT-3’; TF0596-wild , 5’-CCCTGTAGCCCTCTGAACTG-3’; TF0596-31 , 5’-GGGCAGGTTCTGTTTCTCTG-3’ . The primer set Neo_new and TF0596-31 amplifies 247 bp of the targeted allele , while the primer set TF0596-wild and TF0596-31 amplifies 450 bp of the wild allele . Body weight of calves at the age of 4 to 12 weeks and body length at 12 weeks were compared among genotypes by ANOVA . cDNAs were amplified by PCR using PrimeSTAR GXL polymerase ( Takara ) . The PCR primers contained a 15-bp sequence of the 5’-end of the cloning vector or of another primer sequence . The primer sequences are shown in S7 Table . Cloning reactions were performed using In-fusion HD cloning kit ( Clontech Laboratories , Inc . , Mountain View , CA , USA ) . All of the amplified sequences were verified by dideoxy sequencing using BigDye Terminator v . 3 . 1 Cycle Sequencing Kit ( Applied Biosystems ) . Briefly , a cDNA encoding full-length bovine FGD3 was amplified by PCR from the bovine fetus kidney cDNA library [40] and inserted into the pCAGGS expression vector [41] with the addition of the C-terminal HA tag . Ser-83 and Ser-87 of the resultant bFGD3-HA cDNA were replaced with alanines and inserted into the pIRES2-EGFP vector ( Clontech ) . The resulting plasmid was termed bFGD3SA-HA/pIRES2-EGFP . His-171-Cys substitution was created from bFGD3-HA cDNA . The SNP at the initial Met was introduced together with replacing the Ser-83 and Ser-87 residues with alanines and inserted into the pIRES2-EGFP vector; the resulting plasmid was termed bFGD3SA-2ndMCys-HA/pIRES2-EGFP . Myc-tagged Cdc42 [24] was cloned into pcDNA3 . 1/V5-His TOPO TA vector ( Life Technologies , Carlsbad , CA , USA ) to add V5-His tag at the C-terminus . Then , the BstXI-NotI fragments encoding EGFP of bFGD3SA-HA/pIRES2-EGFP and bFGD3SA-2ndMCys-HA/pIRES2-EGFP were replaced with Myc-Cdc42-V5-His , resulting in bFGD3SA-HA/pIRES2-Cdc42V5His and bFGD3SA-2ndMCys-HA/pIRES2-Cdc42V5His , respectively . These plasmids were designed to bicistronically express wild or mutant FGD3 ( SA ) and Cdc42 . NIH3T3 cells were transfected with bFGD3SA-HA/pIRES2-Cdc42V5His or bFGD3SA-2ndMCys-HA/pIRES2-Cdc42V5His using Lipofectamine 2000 ( Life Technologies ) according to the manufacturer’s instructions , followed by the incubation for 48h . The cell lysis and pull-down assays were performed as described previously [24] . HA and V5 tags were detected using Anti-HA-tag HRP-DirectT ( MBL International Corporation , Woburn , MA , USA ) and Anti-V5-HRP ( Life Technologies ) with Amersham ECL Plus Western Blotting Detection Reagents ( GE Healthcare UK Ltd . , Buckinghamshire , UK ) , respectively . Chemiluminescence of the respective protein bands was quantified using ImageQuant LAS 4000 ( GE Healthcare ) . The Japan Wagyu Register Association [42] provided normal growth curves of 10 skeletal measurements and body weight for sires and females and of withers height , chest circumference , and body weight for steers . A clear difference was observed in chest circumference between sires and steers at more than 12 . 5 months of age . Thus , the Japanese Black population consisting of 792 steers from a progeny testing program at the Cattle Breeding Development Institute of Kagoshima Prefecture [17] was used to obtain a normal growth curve for every measurement in steers . Growth models were used according to the Japan Wagyu Register Association [42] . Because the average age of steers at start of the progeny test was approximately 9 months , skeletal measurements of steers were divided into two groups at 10 months of age . For measurements collected from steers that were less than 10 months old , the growth curves for sires provided by The Japan Wagyu Register Association [42] were used as a standard , while the growth curves that we produced were used as a standard for measurements of steers that were more than 10 months old . The animals with at least two measurements of less than 2σ ( S . D . ) were recognized as poorly developed . One exception ( ID = 47 ) , which had one measurement of less than 2σ , was defined as poorly developed because it was diagnosed as such at a livestock hygiene service center . An imbalance between chest depth and width or thurl width and pin-bone width was considered only if the difference between the two measurements was more than one σ . | Livestock are typically subjected to intensive artificial selection for traits of economic value to producers . In spite of this strong selection , some major quantitative trait loci ( QTLs ) for an economically important trait never reach fixation in the population . Several studies have revealed that such QTLs are accompanied with an unfavorable effect on other traits of economic importance , including heritable disease phenotypes . The carcass weight QTL , named CW-3 , was previously identified as one of three major QTL in Japanese Black cattle , and it was found to originate from a specific line that had been maintained in a regional subpopulation . Recent efforts to maintain genetic diversity of the Japanese Black breed have resulted in the widespread use of this line throughout Japan . Half-sib QTL analyses of the elite sires repeatedly detected the CW-3 QTL , while skeletal dysplasia has been found in the descendants . Genomic analyses revealed that skeletal dysplasia is inseparably linked with CW-3 and a functional variant of FGD3 was identified as a positional candidate QTN . Further studies such as creating a genetically modified mouse model will be useful to understand a molecular mechanism of FGD3 to modulate bone development . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| []
| 2015 | Non-synonymous FGD3 Variant as Positional Candidate for Disproportional Tall Stature Accounting for a Carcass Weight QTL (CW-3) and Skeletal Dysplasia in Japanese Black Cattle |
Leptospirosis is a worldwide zoonotic disease and a serious , under-reported public health problem , particularly in rural areas of Tanzania . In the Katavi-Rukwa ecosystem , humans , livestock and wildlife live in close proximity , which exposes them to the risk of a number of zoonotic infectious diseases , including leptospirosis . A cross-sectional epidemiological study was carried out in the Katavi region , South-west Tanzania , to determine the seroprevalence of Leptospira spp in humans , domestic ruminants and wildlife . Blood samples were collected from humans ( n = 267 ) , cattle ( n = 1 , 103 ) , goats ( n = 248 ) , buffaloes ( n = 38 ) , zebra ( n = 2 ) , lions ( n = 2 ) , rodents ( n = 207 ) and shrews ( n = 11 ) . Decanted sera were tested using the Microscopic Agglutination Test ( MAT ) for antibodies against six live serogroups belonging to the Leptospira spp , with a cutoff point of ≥ 1:160 . The prevalence of leptospiral antibodies was 29 . 96% in humans , 30 . 37% in cattle , 8 . 47% in goats , 28 . 95% in buffaloes , 20 . 29% in rodents and 9 . 09% in shrews . Additionally , one of the two samples in lions was seropositive . A significant difference in the prevalence P<0 . 05 was observed between cattle and goats . No significant difference in prevalence was observed with respect to age and sex in humans or any of the sampled animal species . The most prevalent serogroups with antibodies of Leptospira spp were Sejroe , Hebdomadis , Grippotyphosa , Icterohaemorrhagie and Australis , which were detected in humans , cattle , goats and buffaloes; Sejroe and Grippotyphosa , which were detected in a lion; Australis , Icterohaemorrhagie and Grippotyphosa , which were detected in rodents; and Australis , which was detected in shrews . Antibodies to serogroup Ballum were detected only in humans . The results of this study demonstrate that leptospiral antibodies are widely prevalent in humans , livestock and wildlife from the Katavi-Rukwa ecosystem . The disease poses a serious economic and public health threat in the study area . This epidemiological study provides information on circulating serogroups , which will be essential in designing intervention measures to reduce the risk of disease transmission .
Leptospirosis is an emerging/re-emerging , worldwide , contagious , bacterial zoonotic disease that affects all mammals , including humans , livestock and wildlife [1 , 2] . The disease is caused by different serovars of pathogenic species of the genus Leptospira [1 , 2] , which is common in tropical and subtropical regions , wherever environmental conditions favour the survival and transmission of the bacterium [3 , 4] . Leptospirosis was first identified by Weil ( 1886 ) and Inada ( 1916 ) [5] . In the East and Central African regions , the disease was reported three decades ago [6] . The sources of infection for humans and other incidental hosts , such as cattle , pigs , horses , and companion animals , are subclinically infected wild and domestic animals , which are the reservoirs for over 250 known serovars of Leptospira [7] . Rodents are the most important source of infection for humans and animals [8 , 9] . The role of rodents as carriers and the main source of leptospiral infection in human has been investigated in some countries . Moreover , different species of rodents , such as Rattus , R . norvegicus , Mus musculus , Bandicota bengalensis , Bandicota indica and Cricetomys gambianus , are known to carry different pathogenic leptospiral serovars [8 , 9] . Leptospira spp lives for a long time in the kidney tubules of an infected animal host , from where they are excreted through the urine [10] . Humans become infected through either direct contact with the urine or other biological materials from the infected animals or indirect contact with water , soil and vegetation polluted with urine from animals harbouring pathogenic leptospires [11] . Leptospirosis is also an occupational disease affecting veterinarians , abattoir workers , sewer workers and other groups of people whose job exposes them constantly to contaminated materials [12] . A serological assay , the Microscopic Agglutination Test ( MAT ) , is considered as the gold standard for the diagnosis of leptospiral infection [12] . The test is used to detect antibodies against different Leptospiral serovars . Previous reports from Tanzania have indicated that leptospiral infection is widely prevalent in humans , livestock , and rodents in some parts of the country [13 , 5 , 14 , 15 , 16 , 17 , 18 , 7] . However , a study on leptospirosis in the Katavi region has not been conducted , suggesting that the role of animals in the transmission and maintenance of the infection is not well understood . Hence , the objective of this study was to establish the seroepidemiology of Leptospira spp and to identify the most prevalent leptospiral serogroups in humans and animals using the Microscopic Agglutination Test ( MAT ) .
The study was carried out between September 2012 and April 2013 in the Katavi region , southwest Tanzania , which is an agro-pastoral community with a wide range of domestic animals and wildlife . The Katavi region is located approximately 6° 30’S and 31° 30’E . All the districts in the Katavi region , namely Mpanda , Nsimbo , and Mlele , ( Fig . 1 ) were involved in this study . Katavi has a tropical climate , with a rainy season from November to April and a dry season from May to October . During the rainy season , rainfall can be extremely high , with a mean annual rainfall greater than 100 mm . The economic activities of the people in the Katavi region are mainly livestock keeping and small-scale farming . For livestock keeping , Katavi residents practice free-range grazing , and for small-scale farming , they cultivate both food and cash crops . During the dry season , the agro-pastoralist graze animals on crop residues , and thereafter , they shift these animals to distant grazing land , commonly known as grazing camps . Different habitats are selected for trapping rodents , such as , plough fields , tiny bushes around homes , marshy areas for the cultivation of rice and sugar , vegetable gardens , and areas with garbage close to homes and within homes . Katavi contains national parks , such as Katavi National Park , which is composed of seasonally flooded grassland plains , miombo woodlands , small lakes , and swampy wetlands [19] . Wild animals commonly found in the park include African buffaloes ( Syncerus caffer ) , elephants ( Loxondata africana ) , zebras ( Equus burchelli ) , impalas ( Aepyceros melampus ) , giraffes ( Giraffa Camelopardalis ) , elands ( Taurotragus oryx ) , baboons ( Papio anubis ) , hippopotamuses ( Hippopotamus amphibious ) , and predators , such as lions ( Panthera leo ) and other small carnivores [20] . The ethical clearance for conducting this study was granted by the Institutional Review Board of Sokoine University of Agriculture ( SUA/FVM/R . 1/9 ) , Medical Research Coordinating Committee of the National Institute for Medical Research , reference number NIMR/HQ/R8a/Vol . IX/1627 , and the Tanzania Wildlife Research Institute ( TAWIRI ) . Additionally , permission was requested and granted from all local authorities in the study area , including TANAPA and the Local Government Authority . Verbal consents were obtained from all the study participants . To safeguard the wellbeing of animals , this study adhered to Animal Welfare Act [21] , as well as the guidelines adopted from the Australian government [22] . A cross-sectional epidemiological study was carried out , in which a multistage cluster sampling was conducted . Villages were randomly selected as the primary unit , from which a total of 138 households were chosen from the list of agro-pastoralists using random numbers . Members of the selected households were subjected to a random selection to obtain a total of 267 humans who were readily available , regardless of their health status . Our target was households with domestic animals . Thus , using cluster sampling , a total of 1351 apparently healthly livestock were selected from the same households where humans were sampled , as described above [23] . Calves and kids below three months and children below the age of two years were not sampled . Wild animals were also targeted for this study , and a total of 42 of these animals were sampled opportunistically . Rodents and shrews were trapped from the sites located near human settlements and near human activities , including homes and crop stores in open fields where a large number of rodent burrows were observed . A total of 207 live rodents and 11 shrews were captured using Sherman LFA Live traps ( 7 . 5 × 9 . 0 × 23 . 0 cm; HB Sherman Traps , Inc . , Tallahassee , FL ) . The traps were placed inside and surrounding the selected houses early in the evenings , and peanut butter mixed with concentrates was used as bait . The traps were inspected every morning , and the captured rodents were anaesthetized using ether in cotton swabs before taking samples . The sex , species , weight , age , and location of the trappings of the captured rodents and shrews were identified and recorded . The rodents and shrews were grouped into two classes based on age ( juvenile and adult ) , as previously described [24] . Human blood samples were collected from the brachial vein using 5 ml plain vacutainer tubes . Cattle and goats were manually restrained , and blood samples were collected from the jugular vein using 10 ml plain vacutainer tubes . For wild animals , buffaloes were captured by darting using a combination of 5–8 mg etorphine hydrochloride ( M99 9 . 8 mg/ml ) ( Novartis , Kempton Park , South Africa ) and 50–80 mg azaperone tartarate , and zebras were immobilized using a combination of 6–7 mg etorphine hydrochloride ( M99 ) and 80 mg azaperone . Lions were immobilized using a combination of 2 . 5 mg/kg ketamine hydrochloride and 0 . 1 mg/kg medetomidine hydrochloride ( Kyron , Pty , SA ) . The drug was remotely injected using a darting gun . The antidote , diprenorphine hydrochloride ( M5050 ) ( Novartis , Kempton Park , South Africa ) , was used to revive the buffaloes and zebras after the collection of the blood samples . Lions were revived using antisedan ( atipemazole hydrochloride ) . The blood from these animals was collected from the jugular vein using 10 ml plain vacutainer tubes . The blood samples from both domestic and wild animals were allowed to clot in a slanted position , and serum samples were harvested after 24 hours . For rodents and shrews , blood was collected from the retro orbital sinus using sterile capillary tubes and then transferred to eppendorf tubes . The samples were centrifuged , and sera were immediately harvested . The sera harvested from domestic animals , wildlife , rodents and shrews were dispensed into appropriately labelled 1 . 5 ml cryovials and stored in liquid nitrogen ( −78°C ) before being transferred to the Faculty of Veterinary Medicine , Sokoine University of Agriculture laboratories and stored in an ultra-deep freezer ( −80°C ) until a subsequent MAT was performed . Seven leptospiral serogroups , including local isolates , Icterohaemorrhagie ( Leptospira interrogans serovar Sokoine ) , Australis ( Leptospira interrogans serovar Lora ) , Ballum ( Leptospira borgpetersenii serovar Kenya ) and Grippotyphosa ( Leptospira kirschneri serovar Grippotyphosa ) , and reference serogroups , Sejroe ( Leptospira interrogans serovar Hardjo ) , Hebdomadis ( Leptospira santarosai serovar Hebdomadis ) and Canicola ( Leptospira interrogans serovar Canicola ) , which are commonly found in Tanzania , were used in the study . All sera were tested for antibodies against live antigens suspensions of Leptospira spp serogroups Icterohaemorrhagie ( Sokoine ) , Australis ( Lora ) , Ballum ( Kenya ) , Gripotyphosa ( Grippotyphosa ) , Sejroe ( Hardjo ) , Hebdomadis ( Hebdomadis ) and Canicola ( Canicola ) by MAT , as previously described by [25] and [26] . Briefly , the sera ( 10 μl ) were diluted with phosphate buffered saline ( PBS ) to obtain 100 μl of diluted sera in ‘U’ microtitration plates to obtain an initial dilution range of 1:20–1:160 . Then , 50 μl of the full-grown antigens in Ellighausen—mcCoullough/Johnson-Harris ( EMJH ) with an approximate density of 3*108 leptospires/ml on the MacFarland scale was added to all microtiter plate wells and mixed thoroughly on a microshaker . The microtitration plates were then incubated at 30°C for two hours . The serum antigen mixture was visualized under dark field microscopy for the presence of agglutination/clearance , and the titers were then determined . A serum was considered positive if 50% or more of the microorganisms in the microtiter well were agglutinated at the titer ≥ 1: 160 . This was determined by comparing 50% of spirochaetes , which remained free cells with a control culture diluted 1:2 in phosphate-buffered saline [26] . In this study , we examined the positive and negative controls and selected the samples that agglutinated more than halfway through , as previously described by International Committee on Systematic Bacteriology [27] . The samples that agglutinated were identified during the screening of 1:160 dilutions , the numbers were recorded , and the sera were further diluted to determine the end point titer for each sample . The agglutinating sera were tested again at dilutions of 1:20 , 1:40 , 1:80 , 1:160 , 1:320 , 1:640 , 1:1 , 280 , 1:2 , 560 , 1:5 , 120 , 1: 10 , 240 and 1:20 , 480 . Negative and positive controls were included in each test . Phosphate-buffered saline ( PBS ) was used as a negative control . As a negative control , an equal ( 50 μl ) volume of PBS was mixed with the different antigens . The positive control used in this study was rabbit antiserum of each specific serogroup . The positive control antiserum was supplied by the WHO Reference laboratory at the Royal Institute of Hygiene ( KIT ) , Amsterdam , Netherlands . We used seven different positive control antisera from rabbit to test samples from animals and humans , regardless of the species tested . Microsoft Office Excel® 2007 ( Microsoft Corporation , Redmond , 98052-7329 , USA ) was used for storing data and drawing graphs . The prevalence of leptospiral antibodies was computed using Epi-Info version 7 ( CDC Atlanta , USA ) . The proportions were compared using MedCalc® version 13 . 0 . 2 ( MedCalc software , Acacialaan 22 , B-8400 , Ostend , Belgium ) .
The overall prevalences of leptospiral antibodies in human , domestic ruminants , wildlife , rodents and shrews were 29 . 96% , 26 . 35% , 28 . 57% , 20 . 29% and 9 . 09% , respectively . The specific prevalences of leptospiral antibodies in human and in different animal species are indicated in Fig . 2 . One of the two sampled lions was seropositive . Statistical analysis of the results for cattle and goats demonstrated a significant difference in seroprevalence between the two species ( difference 21 . 90% , 95% CI = 16 . 93–26 . 12 , P<0 . 0001 ) . No leptospiral antibodies were detected in zebra ( n = 2 ) . The association of leptospiral infection with sex and age in humans , cattle , goats , rodents , and shrews was not statistically significant ( P > 0 . 05 ) . In this study , leptospira antibodies were detected in 42 ( 20 . 29% ) out of 207 apparently healthy rodents tested . Additionally , 11 shrews were tested , and one ( 9 . 09% ) was found positive . One shrew and seven rodent species were found positive , with varying prevalence among species . The prevalences of leptospira antibodies among different rodent and shrew species are shown in Table 1 . The geographical distribution of leptospiral antibodies were based on the MAT results and are shown in Fig . 1 . The proportions of seropositive individuals exposed to different serogroups , for humans , domestic ruminants and rodents , are presented in Table 2 . Australis was the only serogroup exposed to shrews . In buffaloes , the detected antibodies specific for the serogroups were Sejroe ( 7 . 89% ) , Hebdomadis ( 7 . 89% ) , Australis ( 5 . 26% ) , Grippotyphosa ( 5 . 26% ) , and Icterohaemorrhagie ( 5 . 26% ) , and antibodies against Sejroe and Grippotyphosa were detected in one of the two lions . The results also showed that samples from 13 humans , 63 cattle , two goats , and two rodents reacted to more than one serogroups ( Table 3 ) . In buffaloes , the three positive samples showed serological cross-reactions with two serogroups , specifically , Icterohaemorrhagie and Sejroe , Australis and Grippotyphosa and Hebdomadis and Icterohaemorrhagie . With regards to the lion , the positive sample showed cross-reactions of serogroups Sejroe and Grippotyphosa . The distributions of the different serogroups among the seropositive humans , domestic ruminants , wildlife , rodents and shrews are shown in Fig . 3 .
The findings from this study indicate that leptospiral antibodies are prevalent in the Katavi-Rukwa ecosystem , as the antibodies were detected in humans , cattle , goats , buffaloes , lions , rodents and shrews . This is the first report of leptospira seroprevalence linking humans and animal infections in Tanzania . The demonstration of the exposure of these animals and humans at the same time provides a significant and important epidemiological picture and increases our understanding of infection patterns of leptospiral serogroups at the interface areas . Previous studies demonstrated that the seroprevalence of leptospira in healthy animals suggests levels of local exposure [28] . Animals with low prevalence of leptospira antibodies might be a significant cause of infection in humans , and high seroprevalence may signify exposure pressure from different animals and thus a high infection risk in humans as well [28] . In domestic animals , the highest seroprevalence was observed in cattle , as opposed to goats , and the difference was statistically significant ( P<0 . 0001 ) . The observed difference can be attributed to the feeding behaviour of goats , specifically , grazing on the top end of grasses , browsing on shrubs and staying in less wet areas , as opposed to cattle . As such , they have less exposure to leptospires [29] . In the present study , no significant difference in seroprevalence according to age was established in humans and in other animal species . The study results demonstrate that leptospirosis is endemic in the study area . This implies that all age groups face equal risk of being infected by leptospires . This finding is in agreement with the observation made by other researchers [30] . In humans , the serogroup Sejroe ( serovar Hardjo ) was the predominant serogroup , followed by Icterohaemorrhagie . Other prevalent serogroups were Grippotyphosa , Hebdomadis , Ballum and Australis . The predominance of serogroup Sejroe ( serovar Hardjo ) in humans can be attributed to the high contact rate with cattle , which are widespread in areas where human subjects were sampled for this serosurvey . Cattle are known to be natural hosts for serovar Hardjo , and the spirochete can survive in cattle for years [31] . Interactions between humans and cattle can lead to the interspecies transmission of serovar Hardjo . The seropositivity of serogroup Icterohaemorrhagie ( serovar Sokoine ) and Grippotyphosa in humans can be attributed to the abundance of rodents in the study area , as rodents are the natural carriers of these serogroups [31] . Antibodies to serogroup Ballum ( serovar Kenya ) were detected in humans but not in the sampled animal species in the ecosystem . The serogroup was previously isolated from urine of African giant pouched rat ( Cricetomys gambianus ) from Morogoro , Tanzania [14] . The seropositivity of the serogroup Ballum in humans may be due to the presence of African giant pouched rats in the study area , which may serve as a potential source of the serogroup to humans due to contamination of the environment with urine . The main serogroup identified in cattle was Sejroe ( serovar Hardjo ) . This finding corresponds well with findings in previous published reports that showed that cattle are the maintenance host of this serogroup [5 , 18] . However , studies conducted in different areas of Tanzania have reported significant lower seroprevalence ( 5 . 6% ) than what was observed in the current study [5] . The observed difference in the results between the current and the previous studies is likely due to variations in ecological factors , such as humidity , climate , and environmental factors [32 , 33] , as well as a variation in the level of interaction with other animals in the study area . Serovar hardjo is considered to be an important cause of bovine leptospirosis , which , in most cases , has been associated with abortion in cattle and has also been the most common cause of leptospiral infection in humans , due to the possibility of high rates of interaction between cattle and humans [31] . Other serogroups detected in cattle , such as Hebdomadis , Australis , Grippotyphosa and Icterohaemorrhagie , are accidental infections that are carried by other domestic and free range animals , and which are dependent on farm management practices , as described elsewhere [34] . Icterohaemorrhagie , Sejroe ( serovar Hardjo ) and Grippotyphosa were the most prevalent serogroups observed in goats . Similarly , rodents are known to be the natural reservoir hosts for the serogroups , Icterohaemorrhagie and Grippotyphosa [31] . Therefore , the high prevalence of these two serogroups in goats implies that there is probably high rate of interaction between goats and rodents in the study area . Furthermore , the results suggest interaction between these animals and humans , as the same serogroups were detected in human samples in the same interface ( Table 2 ) . The serogroups Icterohaemorrhagie ( serovar Sokoine ) and Sejroe ( serovar Hardjo ) have previously been reported to be among the most important occupational diseases in and around Tanga city , eastern Tanzania [35] . Antibodies to different leptospiral serogroups were detected in seven different species of rodents ( Table 1 ) trapped in various areas of the Katavi-Rukwa ecosystem , suggesting that rodents are probably the carriers of different leptospiral serogroups , therefore exacerbating transmission of leptospiral infection to humans and animals in the ecosystem . Serogroup Australis had the highest seroprevalence ( 18 . 84% ) , followed by Icterohaemorrhagie ( 1 . 93% ) and Grippotyphosa ( 0 . 48% ) , in the tested rodents in the study area . These findings are in agreement with the findings in previous studies conducted in different parts of Tanzania [11 , 7] . Interactions among rodents , humans , domestic ruminants and wildlife occur frequently in the study area , as rodents share the same habitat with these animals and humans . These interactive activities of rodents in the study area create a favourable environment for leptospiral transmission from rodents to humans , domestic animals and wildlife . Australis was the only serogroup exposed to shrews , with a seroprevalence of 9 . 09% . Exposure to the serogroup was also detected in both rodents and shrews ( insectivores ) . This study was not able to demonstrate whether shrews were the maintenance host for this serogroup or if the serogroup was transmitted to shrews from rodents . This lack of clear understanding of the maintenance hosts may require further studies in this area , as it poses a major public health risk . In buffaloes , the predominant serogroups were Sejroe ( serovar Hardjo ) , Hebdomadis , Australis , Grippotyphosa and Icterohaemorrhagie . The seropositivity against serogroup Serjoe ( serovar Hardjo ) as a predominant serogroup is probably due to high interaction with cattle , which are the maintenance host of the serogroup [36] . Buffaloes are also a reservoir for Hardjo . Hence , the possibility of transmission of the serovar from cattle to the buffaloes and vice versa is very high . The prevalence of serogroup Grippotyphosa and Icterohaemorrhagie in buffaloes may be attributed to the high rate of interaction between buffaloes and rodents in the area because rodents are the carriers of the serogroups [33] . As noted earlier , the leptospiral serogroups found in the sampled buffaloes were similar to the serogroups detected in cattle . The presence of a wide range of buffaloes in Katavi allows cattle and buffaloes to share grazing grounds and watering points . Hence , the possibility of transmission or spillover of serogroups from cattle to buffaloes and vice versa is very high . A similar observation was also reported in Turkey , where researchers found a similar leptospirosis seroprevalence in buffaloes as that observed in cattle [36] . In this study , Sejroe ( serovar Hardjo ) and Grippotyphosa were the only leptospira serogroups detected in lions ( only two lions were sampled and only one was seropositive ) . These same serogroups were also found in buffaloes and domestic ruminants . This may be attributed to the feeding behaviour of lions that prey on wild and domestic ruminants . The current study identified similar leptospiral serogroups circulating in humans , domestic ruminants , wildlife , rodents and shrews sharing the same ecosystem ( Fig . 3 ) . This may be attributed to the intense overlap of these species , bush meat handling , and environmental and seasonal drivers , such as drought and floods [37] . Katavi residents are mainly agro-pastoralists , who frequently come into contact with livestock , as well as wildlife and their excreta , in the ecosystem . Furthermore , the majority of the communities in the study area slaughter animals at home , and some of the people consume raw kidney and liver , as was reported in the interviews conducted . It is believed that direct contact between humans and animals is an important risk factor for human Leptospirosis [38 , 5 , 39 , 40] . The results of this study indicate that livestock share the same serogroup with humans , and this implies a public health risk , particularly among those involved in animal handling . Similar findings were observed in Italy , where patients were infected through direct contact with infected animals or through contaminated urine [41] . In Katavi National Park , wild ungulates are found in high densities around lake Chada , the Kitusunga flood plains , and around lake Katavi , especially during the dry season , due to an influx of wildlife in search of pasture and water [20] . The large influx of animals might easily contaminate the area with urine and increase the chances of the spillover of infections to other animal species . During the rainy season , rivers flood , which increases the risks of leptospirosis outbreaks due to runoff soil contaminated with urine from domestic animals , wildlife , rodents and shrews flowing into common water sources . This is an important driver of leptospiral transmission . In the Katavi-Rukwa ecosystem , bush meat consumption is common , and the main species hunted are impala , common duiker , warthog , buffalo and bushbuck [19] . Leptospiral infection in humans can occur through the direct contact with the blood , tissues , organs and urine of infected animals [39 , 40] . Therefore , slaughtering and handling of bush meat from infected animals may pose a great risk of leptospiral transmission to humans in areas where consumption of bush meat is practiced . A study carried out in South America showed that human leptospirosis was associated with men who captured , slaughtered , and consumed large rodents [42] . This study observed serum agglutination to more than one serogroup in humans and all animal species tested . This may reflect a mixed or two different past infections , which most frequently reflect serological cross-reactions . These cross-reactions are mostly seen in acute or early convalescent sera , whereby the host , previously infected with one serogroup , may subsequently become infected by another serogroup , and the recently acquired serogroup may cross-react to the previous one , leading to activation of the memory response against the subsequent serogroup [43] . The titer of antibodies relating to the previous serogroup could be higher than antibodies specific to the new infecting serogroup . This may also reflect an infection caused by a serogroup not included in the MAT panel , as the MAT panel used was not very wide . In conclusion , the present study demonstrates the possible interaction between livestock , wildlife , and humans , as similar serogroups were detected among these species . This may have a very serious implication on the public health of the communities , as the capacity to diagnose leptospirosis is not available in any of the surveyed villages , including the district hospital . Therefore , human leptospirosis should be included in the differential diagnosis of febrile illnesses in humans in the study area . Furthermore , the results from this study demonstrate common serogroups circulating among humans , domestic ruminants and wildlife , which will help in planning for interventions for the control or mitigation of the impact of infections in domestic animals and in humans . Thus , we recommend further studies on the molecular typing of leptospiral isolates from humans and from different animal species in the Katavi- Rukwa ecosystem . | Leptospirosis is a disease of worldwide significance , and it is also an important zoonotic disease , particularly in developing countries . Subclinically infected rodents maintain leptospires in nature , and some that recover from the primary leptospiral infection may release the bacterium in their urine for the rest of their lives . These rodents serve as a potential source of leptospiral infection to animals and humans . Non-rodent mammals can also be reservoirs of leptospiral infection to animals and humans . Globally , animal and human leptospirosis has been attributed to rodents . There is limited knowledge on the occurrence of the disease in domestic animals , humans , wildlife and rodents in many parts of Tanzania , including the Katavi-Rukwa ecosystem . Serological examination of cattle , goats , humans , buffaloes , zebra , lions , rodents and shrews in the ecosystem revealed the presence of antibodies to serogroups Sejroe , Hebdomadis , Grippotyphosa , Icterohaemorrhagie , Australis and Ballum . These serogroups infect not only their usual hosts but also other animal species , which can in turn act as reservoirs of these serogroups to other animals and humans . This study demonstrates the distribution of leptospiral serogroups in domestic animals , humans , wildlife , rodents and shrews in the Katavi-Rukwa ecosystem . The results of the current study will help in developing appropriate interventions for preventing or mitigating the impacts of infections in domestic animals , humans , wildlife , rodents and shrews . Our results also suggest that human and animal populations are at risk of contracting the infection . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
]
| []
| 2015 | Predominant Leptospiral Serogroups Circulating among Humans, Livestock and Wildlife in Katavi-Rukwa Ecosystem, Tanzania |
The infective schistosome cercaria develops within the intramolluscan daughter sporocyst from an undifferentiated germ ball , during which synthesis of proteins essential for infection occurs . When the aquatic cercaria locates the mammalian host it rapidly penetrates into the epidermis using glandular secretions . It then undergoes metamorphosis into the schistosomulum , including replacement of its tegument surface membranes , a process taking several days before it exits the skin . Patterns of gene expression underlying this transition have been characterised . All gene models from the S . mansoni genome ( www . GeneDB . org ) were incorporated into a high-density oligonucleotide array . Double-stranded cDNA from germ balls , cercariae , and day 3 schistosomula was hybridised to the array without amplification . Statistical analysis was performed using Bioconductor to reveal differentially transcribed loci . Genes were categorised on the basis of biological process , tissue association or molecular function to aid understanding of the complex processes occurring . Genes necessary for DNA replication were enriched only in the germ ball , while those involved in translation were up-regulated in the germ ball and/or day 3 schistosomulum . Different sets of developmental genes were up-regulated at each stage . A large number of genes encoding elastases and invadolysins , and some venom allergen-like proteins were up-regulated in the germ ball , those encoding cysteine and aspartic proteases in the cercaria and schistosomulum . Micro exon genes encoding variant secreted proteins were highly up-regulated in the schistosomulum along with tegument and gut-associated genes , coincident with remodelling of the parasite body . Genes encoding membrane proteins were prominently up-regulated in the cercaria and/or day 3 schistosomulum . Our study highlights an expanded number of transcripts encoding proteins potentially involved in skin invasion . It illuminates the process of metamorphosis into the schistosomulum and highlights the very early activation of gut-associated genes whilst revealing little change in the parasite's energy metabolism or stress responses .
Schistosomiasis mansoni remains an important water-borne disease of humans in Sub-Saharan Africa and parts of South America . Transmission between humans and the aquatic intermediate molluscan host is effected by a miracidium . This free-swimming larva hatches from eggs excreted in the faeces and penetrates into the snail . There follows a phase of asexual multiplication within the snail , before the non-feeding infective cercariae emerge . These have a free life of only hours during which they must locate a human host or perish . Infection occurs when a cercaria penetrates through the skin and transforms into the schistosomulum stage . There is then a period of physiological and morphological adaptation to the new environment , lasting several days , before the parasite locates a blood or lymphatic vessel to exit the skin and begin its intravascular migration to the portal system . The biological processes associated with the transition from snail haemolymph via fresh water to mammalian tissues are the key to an understanding of the infection process . Indeed , it could be argued that preventing parasite establishment would provide the optimum control strategy for the disease . The secretions used by the cercaria to enter the skin have been a focus of interest for decades , latterly using biochemical [1] and proteomic techniques [2] , [3] . These studies have revealed the importance of serine- and metallo-proteases plus potential immunomodulators released by the parasite to gain entry and establish in the skin . However , the relative paucity of parasite material coupled with the limited sensitivity of proteomic techniques means that we have only a partial picture of one small aspect of the transition from snail to human . For example , virtually nothing is known about processes associated with development of the germ balls that mature into cercariae within the daughter sporocyst , nor is the transformation into the schistosomulum well characterised . The cercarial tegument is known to be shed and replaced by the novel double bilayer structure [4] accompanied by the appearance of glucose transporters on the schistosomulum surface , doubtless to facilitate nutrient uptake [5] . The switch from aerobic to anaerobic metabolism has also been noted [6] . The highly sensitive methods now available to characterise gene expression present an opportunity to gain a deeper understanding of this important transition in the life cycle . Microarrays have become a widely used tool for comparing transcription levels between different biological samples; it is good practice to use at least three biological replicates ( i . e . material from separate organisms ) to get statistically significant results for each comparison made . A variety of schistosome arrays has previously been designed and used to answer distinct questions about the parasite's biology . These have included differences in gene transcription between female and male adult worms [7] , [8] , with laser-capture micro-dissection added to facilitate comparison of gastrodermis and vitellaria [9] , and changes in transcription between different life cycle stages [10]–[13] . Germane to our current study , a custom cDNA array , comprising 6000 features from the lung stage larva was used to identify transcripts enriched at the lung stage compared to six other life cycle stages [12] . The same array was also used to characterise differences in transcription pattern between schistosomula transformed from normal and radiation-attenuated cercariae , and cultured in vitro for four , seven and ten days [14] . Other arrays , comprising 12–38 , 000 synthetic oligonucleotide probes have been used to investigate a range of life cycle stages . These arrays were based on ESTs and contigs available pre-2006 at the DCFI S . mansoni Gene Index ( formerly the TIGR Gene Index ) [11] , [15] or ESTs and contigs available at GenBank and the Wellcome Trust Sanger Institute ftp site in May 2005 [13] . Infected snail hepatopancreas , cercariae and adult worms , have been compared using such oligonucleotide arrays , with uninfected snail tissue as a control for the first of these samples [11] . It was reported that the intramolluscan parasite has high levels of transcripts encoding proteins involved in translation and quality control , cell death and ubiquitination . In the cercaria , highly expressed genes were mainly involved in mitochondrial function , enabling the energy production necessary for swimming . However , it was noted that the cercaria was less transcriptionally active than the other stages studied . Transcription levels at 15 distinct points throughout the parasite life cycle have also been compared [13] . Data analysis focussed on three gene families , fucosyl transferases , tetraspanins , and G protein-coupled receptors ( GPCRs ) proffered as potential intervention targets . Finally transcription in 3 hr and 5 day schistosomula , cultured in vitro +/− erythrocytes , was compared with cercariae as the baseline [15] . The most apparent changes were the up-regulation of genes involved in blood feeding , tegument and cytoskeleton development , cell adhesion and stress . Although the annotated S . mansoni genome with standardised nomenclature for predicted gene models was published in 2009 [16] , the above microarray studies have used other nomenclatures and annotations for their constituent ESTs and contigs . This makes specific inter-study comparisons about changes in the transcription of named genes extremely cumbersome and when attempted , points up considerable discrepancies in annotation . We report here the design and use of the most comprehensive microarray platform for S . mansoni to gain insights into infection of the mammalian host . The array was used to probe transcripts from three life cycle stages , intramolluscan germ balls , free-living cercariae , and ‘skin’ schistosomula . We use the term ‘germ ball’ to encompass all stages of embryonic development up to , but excluding , the mature cercaria that comprises approximately 1000 cells [17] , differentiated into tissues and organs . The germ balls were essential as it has been shown that many proteins used by the cercaria for host entry are transcribed and translated during its development in the snail [2] , [18] . They were obtained by microdissection of snails 22–26 days post-infection . Day 3 schistsosomula were chosen because by that point , metamorphosis from the cercaria is nearing completion , they have adapted to life in mammalian tissues and are ready to begin intravascular migration . Such schistosomula , transformed and cultured by the methods we used , are able to mature if transferred into the murine host [19] . They are biologically comparable to ex vivo worms and can be produced in large quantities . Sufficient RNA was obtained from all three life cycle stages for hybridisation to the array without PCR amplification . A comprehensive analysis of greater than two-fold changes in transcription between the life cycle stages is presented .
The procedures involving animals were carried out in accordance with the UK Animals ( Scientific Procedures ) Act 1986 , as authorised on personal and project licences issued by the UK Home Office . The study protocol was approved by the Biology Department Ethical Review Committee at the University of York . All parasite material was from a Puerto Rican isolate of S . mansoni maintained at the University of York by passage through NMRI strain mice and albino Biomphalaria glabrata snails . Developing germ balls from daughter sporocysts were obtained from snails infected with 40 miracidia each and dissected carefully in filter-sterilised 50% PBS ( pH 7 . 4 . ) 22–26 days later , before cercarial maturity . Obvious snail material was removed and freed germ balls at all stages of development were accumulated on ice until use . Cercariae were collected from snails infected with 10 miracidia each . Five weeks after infection the snails were placed in the dark for two days and then illuminated in approximately 10 mls aerated tap water for two hours to induce shedding . The emerging cercariae were gravity-concentrated by cooling on ice for one hour , which prevented swimming . Skin stage schistosomula were obtained by mechanical transformation of cercariae and separation of their bodies which were cultured for three days in vitro as previously described [19] . They were then recovered and washed twice in RPMI before processing [19] . RNA was extracted from the three larval stages by homogenisation in TRIzol ( Invitrogen , Paisley , UK ) at approximately 1 ml per 100 µl tissue . The RNA was extracted as per the manufacturer's instructions , with the addition of DEPC-treated high salt solution ( 0 . 8 M sodium citrate and 1 . 2 M NaCl ) at the isopropanol step , to remove glycoprotein . RNA was isopropanol-precipitated overnight at −80°C with 1 µl Glycoblue ( 15 mg/ml; Ambion ) to aid the process , and visualisation of the pellet . It was recovered by centrifugation at 12 , 000× g for 30 minutes at 4°C . The pellets were washed with 70% ethanol , and allowed to air-dry at room temperature before being resuspended in 300 µl DEPC-treated water . RNA was quantified using a Nanodrop ND-1000 Spectophotometer ( Nanodrop Products Fisher , Wilmington , Delawere , USA ) and quality assessed using a 2100 Bioanalyzer PicoChip ( Agilent , Wokingham , UK ) . The predicted genes from version D of the S . mansoni genome assembly as of June 2008 ( www . GeneDB . org ) formed the input for the array design along with all S . mansoni ESTs available at GeneDB . org whose direction was known , compiled using phrap ( http://www . phrap . org/phredphrapconsed . html ) . The input data were broken up into sequential 50mers offset by one base each time , and redundant sequences were removed using FAlite . pm ( Ian Korf; http://homepage . mac . com/iankorf/ ) and associated Perl scripts . The unique sequences were mapped back to the genome assembly using ‘exonerate’ ( http://www . ebi . ac . uk/~guy/exonerate/ ) . From a map-ordered list , every 13th 50mer was chosen as a probe . No selection was made for the number of probes per predicted transcript . The design was sent to Roche-NimbleGen , who made some minor refinements for ease of synthesis and constructed the arrays using digital micromirror technology [20] . There were 385K features on the array comprising 377 , 598 S . mansoni sequences and 11 , 613 random sequences for hybridisation controls . Double stranded cDNA for hybridisation was synthesised from total RNA using SuperScript Double-Stranded cDNA synthesis kits ( Invitrogen ) according to the protocol supplied by Roche-NimbleGen . The resulting cDNA was pooled such that separate biological replicates were obtained i . e . no parasite homogenates were split across replicates . Roche-NimbleGen were supplied with at least 2 . 7 µg of double stranded cDNA for three biological replicates each from germ balls , cercariae , and day 3 schistosomula to perform the hybridisations . Each biological replicate of cDNA was labelled with Cy3 and hybridised to the array for 16 to 20 hours at 42°C . Slides were washed , and dried before fluorescence data were read using a Roche-NimbleGen MS 2000 Scanner with NimbleScan software . Roche-NimbleGen supplied background-corrected data . All subsequent statistical analysis was carried out using programmes from the Bioconductor suite [21] . The data were quality-assessed by visual inspection of graphical representations of the raw probe level data . Box plots were drawn using the boxplot function from the graphics package . Correlation data were calculated using the cor function from the stats package and heatmaps were made by calling the heatmap . 2 function in the gplots package . All arrays passed the quality assessment . Next , the data were quantile-normalised using the normalizeBetweenArrays function in the limma package . This ensures identical distributions of the data [22] . Following normalisation , the probe level data were summarised to yield ‘gene level’ data . The probes were re-mapped to the S . mansoni gene predictions at www . geneDB . org ( version F ) using ‘exonerate’ . If a probe matched an S . mansoni predicted gene ( Smp ) locus with an e value <1−05 by both nucleotide and protein BLAST , the probe was annotated to that Smp locus . The intensity value for each locus was the mean intensity of all the probes by which it was represented . The resulting gene level data were the input for the differential expression analysis , which was carried out using the limma package . First a linear model fit was performed . This reduced the data for each gene to a mean value from each of the life cycle stages . Next , differential expression data were obtained by performing a contrast analysis . This compares the transcription level of each gene in the following contrasts: Multiple testing was corrected for using the eBayes function which employs the method of Benjamini and Hochberg [23] . This gives an adjusted P value ( adjP ) . Genes which were differentially expressed above a two fold cut-off between any two stages with an B>3 were chosen for further analysis . The B value is the log ( odds that a gene is differentially expressed ) . For example if B = 3 , the odds that a gene is expressed is e3 = 20 , or 1 in 20 , corresponding to a probability of 95% . Log2 quantile-normalised probe level data from the array are deposited at the public database Gene Expression Omnibus ( http://www . ncbi . nlm . nih . gov/geo ) under accession numbers GSE22037 and GPL10466 . For ease of visualisation , statistically significant expression values were converted to relative fold change , with the stage having the lowest expression for each gene set to unity . This facilitates analysis of changes in gene expression associated with the infection process . The actual baseline expression values may be different for each gene , and comparisons of absolute expression levels between genes , based on the charts and tables shown here , are not valid . However , the patterns of gene expression may be compared between genes . Where a data bar is missing for a particular life cycle stage , this is because either no statistically significant gene expression was observed for that life cycle stage , or there was a less than two-fold difference . Finally a hypergeometric test was carried out using the Category package to discover whether particular GO terms were over-represented in the differentially expressed genes compared to the ‘gene universe’ of Smp gene models on the array . As an aid to understanding biological processes associated with the transition of the parasite from the snail , via fresh water to the mammalian host tissues , we grouped the S . mansoni genes into useful categories . Some of these ( DNA replication , translation , energy metabolism , lipid metabolism , and membrane ) were based purely on the classification of genes by Gene Ontology ( http://www . geneontology . org/ ) . In creating a muscle tissue category , we used the GO term ‘actin binding’ . Two of the tissue categories , ‘tegument’ and ‘alimentary tract’ were created using our proteomic analyses [24]–[28] . In the case of ‘alimentary tract’ , as four proteins containing saposin domains had already been identified in worm vomitus , we searched the genome database for other genes encoding saposins and added them to this category . The custom category ‘defence against stress’ was compiled from the literature on schistosomes [29]–[31] . The VAL and MEG categories were based on recently published compilations [32] , [33] . Genes encoding proteins involved in lipid synthesis were culled from supplementary table 11 of the genome paper [16] . The lists of proteases classified by catalytic type ( http://merops . sanger . ac . uk/ ) were abstracted from the S . mansoni genome supplementary table 18 [16] . The ‘development’ category comprised the neural development and TGFβ signalling genes highlighted in the genome paper ( supplementary tables 9 and 15 , respectively [16] ) along with those genes annotated to the GO term ‘Wnt receptor signaling pathway’ while the protein glycosylation category was created by assiduous interrogation of the S . mansoni genome database . RNA from each of the biological replicates was reserved for qPCR to validate the microarray results . Total RNA ( 0 . 413 µg ) from each sample was reverse-transcribed using Superscript II ( Invitrogen ) and primed with oligo-dT according to the manufacturer's instructions , in a reaction volume of 20 µl . After the reaction , the volume was made up to 100 µl with DEPC-treated water . Relative quantitation was carried out using SYBR green PCR Master Mix ( Applied Biosystems ) with 1 µl of cDNA per 25 µl reaction . The PCR was carried out using an ABI7300 ( Applied Biosystems ) according to the manufacturer's instructions with 18S ribosomal RNA [14] as the endogenous control , run for each sample on each plate . Genes chosen and their primer pairs are detailed in Table S1 , with primers designed using PrimerExpress ( Applied Biosystems ) . There were three technical replicates per biological replicate for each gene . Relative quantification was calculated by the ΔΔCT method using Applied Biosystems Sequence Detection Software version 1 . 2 . 3 7000 system .
This term encompasses the basic functions that might be expected to occur in the ‘average’ schistosome cell . This category encompasses the mechanisms involved in the development and differentiation of tissues and the transcription of genes linked to three specific tissues , namely muscle , gut and tegument . This category encompasses groups of genes either with a similar organization ( the MEGs ) or those that encode proteins with a related molecular structure or function . In this paper we have described in detail the changes in gene transcription throughout the mammalian infection process . Evidence for DNA replication and cell division was only seen in the germ ball , while each stage up-regulated different genes involved in development and morphogenesis , including cell-cell adhesion . Thus , neural patterning genes were enriched in the germ ball , while there was considerable activation of genes involved in nerve function in the schistosomulum , coincident with body remodelling . Forward planning was seen throughout , with genes encoding skin penetration enzymes transcribed in the germ ball , and those involved in transformation , gut activity and tegument replacement beginning their up-regulation in the cercaria . Our data suggest there are vastly more proteases involved in skin penetration than hitherto envisaged . Likewise the number of possible gut proteases and transporters is expanded . The up-regulation of many receptors in the schistosomulum raises the thorny issue of whether they interact with host or endogenous ligands . Surprisingly , few stress-related genes were up-regulated in the schistosomulum and , apart from the cercarial enrichment of genes encoding proteins involved in aerobic respiration , there was little change in energy metabolism during the transitions studied . The inscrutable VALs are deployed early in infection , whilst the majority of the MEGS , also of unknown function , are transcribed later with extremely high fold changes . Many of the genes identified throughout our study warrant detailed investigation . | The schistosome cercaria develops from undifferentiated germ balls within the daughter sporocyst located in the hepatopancreas of its snail intermediate host . This is where the proteins it uses to infect humans are synthesised . After a brief free life in fresh water , if the cercaria locates a host , it infects by direct penetration through the skin . It then transforms into the schistosomulum stage , adapted for life in human tissues . We have designed a large scale array comprising probes representing all known schistosome genes and used it in hybridisation experiments to establish which genes are turned on or off in the parasite during these stages in its life cycle . Genes encoding proteins involved in cell division were prominent in the germ ball along with those for proteases and potential immunomodulators , deployed during skin penetration . The non-feeding cercaria was the least active at synthesising proteins . Conversion to the schistosomulum was accompanied by transcription of genes involved in body remodeling , including production of a new outer surface , and gut activation long before ingestion of red blood cells begins . Our data help us to understand better the proteins deployed to achieve infection , and subsequent adaptations necessary for establishment of the parasite in the human host . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"Discussion"
]
| [
"medicine",
"infectious",
"diseases",
"schistosomiasis",
"functional",
"genomics",
"gene",
"expression",
"genetics",
"biology",
"genomics",
"parasitic",
"diseases",
"genetics",
"and",
"genomics",
"dna",
"transcription"
]
| 2011 | Gene Expression Patterns in Larval Schistosoma mansoni Associated with Infection of the Mammalian Host |
How does the growth of a virus depend on the linear arrangement of genes in its genome ? Answering this question may enhance our basic understanding of virus evolution and advance applications of viruses as live attenuated vaccines , gene-therapy vectors , or anti-tumor therapeutics . We used a mathematical model for vesicular stomatitis virus ( VSV ) , a prototype RNA virus that encodes five genes ( N-P-M-G-L ) , to simulate the intracellular growth of all 120 possible gene-order variants . Simulated yields of virus infection varied by 6 , 000-fold and were found to be most sensitive to gene-order permutations that increased levels of the L gene transcript or reduced levels of the N gene transcript , the lowest and highest expressed genes of the wild-type virus , respectively . Effects of gene order on virus growth also depended upon the host-cell environment , reflecting different resources for protein synthesis and different cell susceptibilities to infection . Moreover , by computationally deleting intergenic attenuations , which define a key mechanism of transcriptional regulation in VSV , the variation in growth associated with the 120 gene-order variants was drastically narrowed from 6 , 000- to 20-fold , and many variants produced higher progeny yields than wild-type . These results suggest that regulation by intergenic attenuation preceded or co-evolved with the fixation of the wild type gene order in the evolution of VSV . In summary , our models have begun to reveal how gene functions , gene regulation , and genomic organization of viruses interact with their host environments to define processes of viral growth and evolution .
The gene orders in the genomes of individual negative-sense single-stranded RNA viruses have been conserved [1]–[3] . More specifically , most viruses in the order Mononegavirales , abbreviated here as ( – ) ssRNA viruses , share a similar genome organization: 3′-cap-phos-mat-env-pol-5′ , where cap encodes nucleocapsid protein ( N ) , phos encodes phosphoprotein ( P ) , mat encodes matrix protein ( M ) , env or multiple analogous genes encode envelope protein ( s ) ( G ) or attachment ( H and HN ) and fusion proteins ( F ) , and pol encodes polymerase protein ( L ) ( Figure 1 ) [1] , [2] . It has long been hypothesized that such gene-order conservation and similarity either reflect the absence of a genome recombination mechanism for this virus family [4] or arise from relevant fitness benefits . However , such a hypothesis has been recently challenged by several studies of ( – ) ssRNA viruses . First , a phylogenetic analysis of nucleoprotein and glycoprotein gene sequences of ebolaviruses from natural isolates suggested that recombination between different groups of ebolaviruses had occurred [5] . Another phylogenetic analysis of several genes of Hantaan virus , Mumps virus and Newcastle disease virus also strongly suggested that recombination in ( – ) ssRNA viruses could take place at low rates [6] . In addition , inverted gene orders of Pneumoviruses with similarities in protein and mRNA sequences ( Turkey rhinotracheitis virus ( TRTV ) : 3′-F-M2-SH-G-5′ , respiratory syncytial virus ( RSV ) and pneumonia virus of mice ( PVM ) : 3′-SH-G-F-M2-5′ , avian pneumovirus ( APV ) : 3′-F-M2-SH-G-5′ ) suggest the possibility for recombination events during their evolution [7]–[9] . Second , changes of gene orders in viral genomes have increased replication rates of some ( – ) ssRNA viruses . For example , when F and G genes were moved into promoter-proximal positions , replication rates of RSV mutants were increased up to 10-fold relative to wild type [10] . In addition , shuffling the P , M , and G genes of vesicular stomatitis virus ( VSV ) created mutants that could grow as well or better than wild type [11] . What is then the origin of gene orders in ( – ) ssRNA virus genomes ? If a recombination mechanism was not available , how might the present specific gene orders have been selected from numerous possibilities ? If genome recombination was possible , do the gene orders of current wild-type viruses represent those with the highest fitness ? Answers to these questions could shed light on how RNA viruses have evolved , but they remain challenging to address . To obtain some initial clues , we sought to compare the fitness of all possible gene-shuffled variants of a prototype ( – ) ssRNA virus based on predictions of their growth dynamics . By using mathematical expressions to account for the dynamics of gene expression and known interactions among gene products one may represent the development of virus growth with a mechanistic model of moderate , but not overwhelming complexity . Growth models of viruses aim to account for the synthesis , interactions and degradation of viral intermediates toward progeny production as they utilize the resources of host cells [12]–[14] . Such models can show how genome-wide regulation of viral gene expression can contribute to the integrated development of virus progeny . Previous work has shown how relocation of the gene encoding the bacteriophage T7 RNA polymerase can influence the phage growth [15] . However , the phage T7 genome encodes 56 genes , from which 56 ! ( = 1074 ) linear gene order permutations could be defined , so only a vanishingly small fraction of the total genome-design space could be examined by wet-lab experiments or computer simulations . Here we consider a relatively simple prototype of the ( – ) ssRNA viruses , VSV , which has been widely studied and well characterized [3] , [16] . As shown in Figure 2A , VSV encodes five genes ( N , P , M , G , and L ) , and these genes define 120 gene-order permutations . The five VSV genes play well-established roles in the growth of VSV , as summarized in Figure 2B . Very briefly , the entering negative-sense RNA genome is transcribed from its 3′ single promoter ( called leader region ( Le ) ) by the virion-associated VSV polymerase ( proteins P and L ) . A controlled attenuation of transcription occurs in each intergenetic region , where a fraction of elongating polymerases are released from the genomic templates , producing mRNA levels that progressively decrease from N to L . Specifically , at the Le-N , N-P , P-M , M-G , and G-L junctions 0 , 25 , 25 , 25 , and 95 percent of polymerases entering the junction are respectively released from the templates without transcribing any downstream genes [3] , [11] , [17] . Hence , the relative expression level of any gene depends on its position within the genome; moving genes toward the 3′ or 5′ end of the genome respectively increases or reduces their level of expression . As N protein accumulates , it associates with nascent viral RNAs , creating an RNA-protein ( or “nucleocapsid” ) complex that enables the elongating polymerase to bypass transcription attenuation signals at intergenic regions , causing a switch from transcription to genome replication . Further , as M proteins accumulate , they associate with and condense the genomic nucleocapsid , diverting it away from transcription and replication processes , while directing it toward the formation of progeny virus particles . Finally , particle budding from the cellular membrane incorporates protein G ( not shown ) into the surface of progeny viruses . Unlike viral transcription and replication , the synthesis of viral proteins relies mainly on host translation resources , whose availability can vary depending on the host cell type and the extent to which they are susceptible to infection-mediated inhibition of protein synthesis . In previous work we developed and employed a mathematical model to simulate and analyze the life cycle of VSV [13] . The model accounted for the core regulatory mechanisms of VSV and incorporated available quantitative knowledge on interactions among viral and cellular components during infection . Model predictions for the growth dynamics of several gene-rearranged VSV variants qualitatively matched the experimentally observed growth ranking and gene expression patterns of the variants [13] . These results suggest that the model might be useful for gaining insights into the growth of other gene-permuted variants . Advances in reverse genetics systems and synthetic biology approaches have facilitated the construction of several genome-engineered virus mutants [18] , but the generation of 120 gene-order permuted variants and experimental comparison of their life cycles remains a daunting task . Instead , we employed our mathematical model here to simulate and study how all gene-order permutations of the VSV genome would be predicted to influence its growth .
VSV has five genes in its genome in the order of 3′-N-P-M-G-L-5′ . We generated in silico 119 all possible gene-permuted VSV mutants by keeping the wild type extents of transcriptional attenuation for the first to the fifth between-gene junctions ( e . g . , 0% , 25% , 25% , 25% , and 95% , respectively ) . Using our model we predicted the dynamics of their growth in baby hamster kidney ( BHK ) cells , and then we compared the dynamics with that of wild type virus ( Figure 3 ) . Here , the growth of wild type is the result of previous fitting of our model to experimental data [13] . In addition , parameters in the model were also constrained so simulated variants would satisfy the experimentally observed growth ranking of five mutants having gene orders 3′-N-n-n-n-L-5′ , drawing n from P , M , and G [13] . Due to the existing attenuation mechanism , the gene-order shuffling yielded a large variation in the production of progeny virus particles . Depending on the gene order viable VSV variants produced from 1 to 6 , 000 progeny particles in an infected BHK cell ( Figure 3 ) . However , forty percent of the variants could not produce any progeny at all . Virion assembly started at around two hours post infection for wild type [16] , but the timing was significantly retarded for most mutants ( Figure 3 ) . Although some variants showed faster growth patterns in the early infection stage between 2 . 5 and 5 . 5 hours post infection , the wild type virus overall grew better than most other variants ( Figure 3 ) . Only two mutants having the gene orders 3′-N-M-P-G-L-5′ and 3′-N-M-G-P-L-5′ produced more progeny particles than wild type . To correlate the genome organization of each variant with its fitness , we first divided the 120 variants into five 24-variant groups , where all members of a group had a specific gene at a specific location . For example , all members of the N1 group have gene N in position 1 , and the other positions are defined by the remaining 24 permutations of the four remaining genes . We then calculated the mean and standard deviation of the progeny virion production of the 24 variants in each group ( Figure 4 ) . Our analysis showed that for better viral growth , N gene , whose product is needed in a large quantity for genome encapsidation [2] , should be located toward the 3′ promoter of the genome ( Figure 4A ) , while L gene , whose product is needed in a low quantity for transcription and replication reactions , should be located toward the last position at the 5′ end of the genome ( Figure 4B ) . Specifically , the variants of the L5 group grew much better than the variants of the other four groups ( L1∼L4 ) ( Figure 4B ) , highlighting the importance of minimal expression of L protein for viral growth . This model prediction is consistent with the experimental results that N-gene rearranged VSV variants grow better as N gene is located on earlier positions [19] and overexpression of L protein inhibits the virus growth [20] . In general , structural proteins are in a greater demand than enzymatic proteins during the viral infection cycle . However , the large variations in the virion productions of the N1 and L5 groups ( Figure 4A and 4B ) suggested that neither the assignment of N gene to the first genome position nor the assignment of L gene to the last position is a sufficient condition for optimal virus growth . The virion production gradually drops as P gene is moved toward 3′-proximal positions ( Figure 4C ) . This is tightly coupled with the low composition stoichiometries of P protein in a VSV particle ( Table 1 ) and in a polymerase complex with L protein . Moving P gene to earlier 3′-proximal genome positions will also reduces the expression of other genes whose products are needed in larger amounts . Due to the high composition stoichiometries of N , M and G proteins in a VSV particle ( Table 1 ) , when one of the three genes is located on the last genome position , the viral growth was severely reduced ( Figure 4A , 4D , and 4E ) . The roles of M protein in condensing the genomic nucleocapsids and inhibiting host transcription additionally require a minimum level of M expression , putting an additional constraint that gene M avoid the last position . However , the location of M gene at any of the other fours positions did not strongly affect the viral growth ( Figure 4D ) . Our analysis of 120 , 414 ranking vectors from the predicted growths of the 120 variants led to a more quantitative and systematic understanding of the effects of genome organization on the viral growth . First , the averaged ranking vector , [N , P , M , G , L]BHK = [1 . 87 , 3 . 57 , 2 . 57 , 2 . 90 , 4 . 09] re-emphasizes that for better virus growth N and L genes need to be located on the first and the last genome positions , respectively . The large difference between the rankings of N and L genes ( 4 . 09−1 . 87 = 2 . 22 ) quantifies how important such a genome position separation of the two genes is for the viral growth . The voting results from progeny virions indicate that the gene order , 3′-N-M-G-P-L-5′ , is the most common form to which genome organizations of many progeny virions match more closely than to any other gene order . This further implies that moderate alterations from this identified gene order would likely less perturb virion production compared to alterations from any other gene order . Therefore , the gene order , 3′-N-M-G-P-L-5′ , can be considered a robust form of genome organization . Our second-order analysis using a Pairs matrix , where a component ( i , j ) quantifies to what extent gene i ( listed in the first column ) is preferred to gene j ( listed in the first row ) for an earlier genome position ( Table 2 ) , reinforced our previous results: preferences for genes in early positions start with N , and are followed by M , G , P , and L . Experiments showed that several gene-shuffled VSV mutants can grow like or better than wild type [11] . Those mutants had the gene orders 3′-N-M-P-G-L-5′ and 3′-N-M-G-P-L-5′ . Our previous model fitting results also suggested that increasing the VSV growth rate by gene rearrangement is feasible based on the given VSV regulatory circuit [13] . From the conventional hypothesis that wild type has the most evolved form of genome organization , results of others' study and our simulations raise clear questions: Why is wild type not the fittest ? Could the fitter variants still grow better than wild type in many different cell types ? Can any other variants grow better than wild type in some cell types ? How does gene order systematically affect VSV growth in different cell types ? To obtain some clues we compared in silico the growth of the 120 variants in BHK cells with their growth in delayed brain tumor ( DBT ) cells . Our previous model fitting to the experimental growth of wild type VSV in DBT cells suggested that resources of BHK cells for translation were 6 fold richer but 1 . 4 fold less stable compared to those of DBT cells [13] . Because host factors are mainly involved in VSV translation rather than in transcription and replication [16] , such features relevant to translation would be the most representative basis to distinguish each host cell type as a different supporting environment for VSV growth . Out of the 120 gene-shuffled variants , the wild type grows second in DBT cells ( third in BHK cells ) , which suggests that the wild type gene order might be a slightly sub-optimal or near-optimal product of natural selection ( Figure 5 ) . However , the existence of a mutant ( 3′-N-M-P-G-L-5′ ) whose simulated infection is more productive than wild type in different hosts , raises questions on the origin of the wild-type genome organization . The virion production ranking of the VSV variants for BHK cells is roughly maintained for DBT cells as shown by the upper-left to lower-right diagonal pattern of variant growth rankings ( Figure 5 ) . Specifically , the high fitness rankers ( ≤18th ) for BHK cells also grow better than other variants in DBT cells . The maintained fitness benefits from these specific gene orders even in the significantly different environments imply that such benefits likely arise from enhanced efficiencies of intrinsic viral regulatory mechanisms by the specific genome organizations , rather than from altered virus-host interaction patterns . However , mutants ranked between 19th and 62th , 73th and 120th showed moderate variations in their relative virion productions depending on the host cell type ( Figure 5 ) . This indicates that the extent of virus fitness change by its gene order permutation also depends on the availability and stability of host factors that vary over host cell types . In addition , we also used two types of metrics , Averaged rankings and Pairs , to compare the importance of gene order in the two cell types . First , the averaged ranking of each gene indicates the position in the genome where the gene needs to be located for productive viral growth . Second , as the component Kij in Pairs , corresponding to gene i and gene j , is closer to 1 and 0 , gene i is more and less preferable , respectively , for an earlier genome position compared to gene j ( See the Method section ) . The larger ranking difference in DBT cells between N and L genes ( 4 . 23−1 . 59 = 2 . 64 ) in the averaged ranking vector , [N , P , M , G , L]DBT = [1 . 59 , 3 . 48 , 2 . 72 , 2 . 97 , 4 . 23] , compared to the case of BHK cells ( 2 . 22 ) , showed that locating N and L genes to the first and the last genome positions is more important for viral growth in DBT cells . However , the smaller standard deviation ( SD ) of the rankings of the three other genes , M , P , and G ( SD of 2 . 72 , 3 . 48 , and 2 . 97 = 0 . 39 ( DBT ) vs . 0 . 51 ( BHK ) ) revealed reduced importance of their genome positions compared to the case of BHK cells . The increased difference between rankings of N and L genes for DBT cells compared to the case of BHK cells ( 2 . 64−2 . 22 = 0 . 42 ) is equivalent to the standard deviations of the rankings of other three genes ( 0 . 39 and 0 . 51 ) . This indicates that such host effects are at a level equivalent to the effect of relative positions of M , P , and G genes on viral growth . In addition , the values of the Pairs Kij for i = N and L are closer to 1 and 0 , respectively , than the case of BHK cells , and the values of the Pairs Kij for i = P , M , or G and j = P , M , or G ( Table 3 ) are all closer to 0 . 5 ( Table 2 ) , highlighting increased importance of the genome positions of N and L genes . Therefore , relative genome positions of genes affect the viral growth to a different extent depending on the type of host cells . The role of gene order in controlling the relative protein expression level of ( – ) ssRNA viruses is tightly linked to the partial transcription termination mechanism by which genes more proximal to the 3′ promoter region are favored for transcription . Such an attenuation mechanism would be obtained through evolution to satisfy different degrees of need of each protein during viral infection cycles . If no attenuation mechanism is available , would the wild-type gene order be more productive than all other gene-order mutants ? If so , one might speculate that the wild type gene order was fixed before VSV adopted the current attenuation mechanism . To test this hypothesis in silico , we predicted the growth of the 120 variants again in the absence of an attenuation mechanism . If a polymerase starts transcription from the 3′ promoter region in the absence of attenuation , then it will move through the whole genome , ultimately synthesizing all the five gene transcripts without being released from the genomic template . Even without any attenuation mechanism , all the gene-shuffled variants can produce progeny particles in BHK cells ( Figure 6 ) , a sharp contrast to the case of the wild type attenuation showing that 40 percent of variants produced no progeny ( Figure 3 ) . However , the variation of virion production by gene-order shuffling is just 20 fold ( Figure 6 ) , which is significantly smaller compared to that of the wild type attenuation case ( more than three log variation as shown in Figure 3 ) . Interestingly , even without an attenuation mechanism , some gene orders are more advantageous than others for viral growth . Due to the time required for polymerase to move from one end of the genome to the other end , there are still spatial polymerase concentration gradients on the genomic templates as the intracellular concentrations of L and P proteins fluctuate during infection . These spatial concentration gradients are smaller in the absence of active attenuation , but they are sufficient to generate differences in the levels of different viral mRNAs over time in a gene-order dependent manner . Thus , some VSV variants are fitter than others , even in the absence of attenuation . In the absence of regulated gene expression our variant grouping analysis also showed significantly reduced importance of genome positions of each gene ( Figure 7 ) . For example , the N1 group produces only 4 . 5 fold more virion particles on the average than the N5 group ( Figure 7A ) , a much smaller difference compared to 2900 fold in the presence of the wild type attenuation ( Figure 4A ) . Interestingly , the L3 and L4 groups can produce more virion particles than the L5 group to which wild type virus belongs ( Figure 7B ) . This suggests that in the absence of attenuation L protein expression does not have to be minimized for viral growth . Further , the N5 and M5 groups showed significant levels of virion production , which is also in contrast to the case of the wild type attenuation ( Figure 4A and 4D versus Figure 7A and 7D ) . The averaged ranking vector , [N , P , M , G , L]BHK-noATT = [2 . 27 , 3 . 62 , 2 . 67 , 3 . 13 , 3 . 31] , highlights a reduced importance of gene order . For example , the ranking gap between the first and the last rankers is only 1 . 35 ( = 3 . 62−2 . 27 ) , significantly reduced from 2 . 22 in the case of attenuation . Despite the reduced effect of gene order , the first genome position is still strongly preferred by N gene for viral growth . The fitter VSV variants in the presence of the wild-type attenuation mechanism tend also to be fitter in the absence of attenuation , as indicated by the upper-left to lower-right diagonal pattern of growth rankings ( Figure 8 ) . However , large deviations from the diagonal indicate a lack of strong correlation . Without an attenuation mechanism , wild type still grows well in BHK cells , but just within the top 24 percent of the 120 variants , compared to its ranking in the top 2 percent in the presence of the wild type attenuation mechanism ( Figures 5 and 8 ) . In addition , the mutant having the gene order 3′-N-M-P-G-L-5′ , which was a fitter compared to wild type in the presence of attenuation , still grows better than wild type in BHK cells . The large number of mutants that are predicted to grow better than wild type in the absence of attenuation has evolutionary implications . Specifically , if natural selection was critical for the fixation and conservation of the wild-type gene order , then the attenuation mechanism should have preceded or co-evolved with gene order . We have shown that the wild type ranks highly under the attenuation mechanism ( Figures 3 and 5 ) . However , the existence of a few mutants that grow like wild type or better could not be clearly explained by our simulations . Several factors may be relevant . BHK and DBT cells may significantly differ from the cell types that VSV infects in nature . Our model may well still lack information on unknown functions of viral proteins or their interactions with cellular components that affect growth . We finally suggest another reason why the sub-optimal genome organization for viral growth was fixed from VSV evolution . Instead of gene-rearrangement requiring a series of complicated recombination steps , there could have been an alternative mechanism to increase the viral fitness by fine-tuning the relative gene expression level . In this setting transcriptional attenuation would be a plausible mechanism . We have already shown the importance of the attenuation mechanism for viral growth under limited host resources . The maximum and the average burst sizes of the 120 variants for BHK cells were increased by 5 . 8 and 4 . 0 fold , respectively , in the presence of the wild type attenuation mechanism compared to the case of no attenuation ( Figures 3 and 6 ) . In particular , the growth of wild type was increased by 16 . 3 fold in the presence of attenuation . Further , our simulations showed that changes of extents of attenuation at gene junctions can increase the virion production of wild type in BHK cells by 24 percent ( unpublished data ) . In addition , depending upon the attenuation pattern virion production from the wild type gene order could vary by 6700 fold ( unpublished data ) . From these predicted large variations of growth phenotype by mutations to the attenuation mechanism , we conjecture that perturbations of the degrees of attenuation at each intergenic region by point mutations could have provided a means for VSV to more readily adapt to new host environments than by re-ordering the wild type genome . This idea is in part supported by experimental observation showing that a few point mutations at an intergenic region of VSV could cause transcriptional attenuation to span from 5 to 98 percent [21] . The control of relative level of viral transcripts is the central mode of regulation during VSV infection cycle . We suggest that VSV has obtained a near-optimal transcription control by co-evolution of its gene order and intergenic sequences instead of relying only upon gene-order optimization for growth .
To consider in detail the transcription attenuation mechanism of VSV , we modeled the spatial and temporal changes of polymerases distributed along the viral genome templates during our simulations of the viral infection cycle [13] . We first partitioned the genomic templates into multiple segments , then estimated the polymerase flux into each segment at each time point post infection , and finally correlated the polymerase occupancy on a fixed number of segments corresponding to each gene with its temporal transcription level [13] . By changing the gene scanning order of polymerase in silico , our model could be easily extended to predict the growth dynamics of gene-shuffled VSV variants . While the gene order of each variant affects its transcription pattern , we assumed the intrinsic interactions among encoded viral proteins and RNAs to be conserved among all variants . Using our model we simulated the cell infections by each of the 120 gene-shuffled VSV variants and determined the resulting yield of virus progeny . For example , two virus variants having the gene orders 3′-G-L-M-N-P-5′ and 3′-L-M-G-P-N-5′ produced 645 and 1 virion particles in an infected BHK cell , respectively . Based on their virion production , they were ranked as 48th and 80th , respectively . By grouping the 120 VSV variants based on the genome position of a specific gene and comparing the averaged progeny production of each group , we quantified how increasing or decreasing the relative expression level of a single gene affects the virus growth . For example , based on the location of N gene five groups can be defined ( e . g . , 3′-N-n-n-n-n-5′ ( N1 ) , 3′-n-N-n-n-n-5′ ( N2 ) , through ( N5 ) , where n is either P , M , G , or L ) . Each group consists of 24 virus variants that contribute to the calculation of the average ( mean ) and standard deviation of virus production for the group . To better understand how relative gene order impacts progeny production we viewed the simulated virus growth as voting results . The 120 VSV variants produced a total of 120 , 414 virion particles in individually infected BHK cells . Now we assume that each virion particle as a voter ranks five different candidates ( N , P , M , G , and L ) . For example , 645 virion particles ( having the gene order , 3′-G-L-M-N-P-5′ ) choose G as the first ranker , L as the second , and so on . From this voting result we can construct 645 ranking vectors for these 645 virion voters ( y1 , y2 , … y645 = [4] , [5] , [3] , [1] , [2]T ) . In each ranking vector we put the rankings of N , P , M , G , and L , first to fifth , respectively . In this manner we generated 120 , 414 ranking vectors for the total 120 , 414 virion particles . Two metrics calculated from such ranking vectors systematically quantified the impacts of the location of each gene as well as interactions among locations of different genes . Averaged ranking [22] ( 1 ) where yi is a ranking vector , n is the number of total virion particles as voters , and Y is the averaged ranking vector for the virion voter population . Depending on the type of host cell that interacts with viruses and has a certain level of resources for biosynthesis , n will vary . The averaged ranking can inform where a single gene needs to be located within the genome for productive viral growth . For example , from Y = [3 . 21 , 1 . 32 , 4 . 33 , 2 . 21 , 4 . 67]T we can infer that for viral growth N , P , M , G and L genes need to be located at the third , first , fourth , second , and fifth genome positions , respectively . Pairs [22] ( 2 ) where if the ranking of component i is higher than the ranking of component j for the ranking vector k , then count one . After the counting process for all the ranking vectors , the obtained numbers are summed ( # ) and then divided by the number of total virion voters ( n ) . K is the Pairs matrix ( 5×5 for our case ) that shows which gene should be followed by which gene for productive viral growth . As a numerical example , we assume that we have three ranking vectors ( or voting results ) , [2] , [4] , [1] , [3] , [5]T , [2] , [4] , [5] , [1] , [3]T and [3] , [1] , [5] , [4] , [2]T , and we consider component 1 and 2 . For K12 we obtain 2/3 by counting 2 from the first two ranking vectors and then by dividing it by 3 ( = n ) . As the number corresponding to gene i and gene j ( Kij ) is closer to one , gene i is more preferable for an earlier genome position compared to gene j . In contrast , as the number is closer to 0 , then gene i is less preferable to gene j . If the number is close to 0 . 5 , then the relative genome positions of gene i and j are not important for the virus growth . | Although many viruses are linked to diseases that adversely impact the health of their human , animal , and plant hosts , viruses could help promote wellness and treat disease if their “good traits” could be harnessed . Potentially useful virus traits include their abilities to stimulate a robust immune response , target specific tissues for the delivery of foreign genes , and destroy tumors . The exploitation of such traits in the engineering of virus-based vaccines , gene therapies and anti-cancer strategies is limited in part by our inability to control how viruses grow . Generally , viruses that grow poorly will be more desirable for vaccine applications , whereas viruses that grow and spread rapidly will be useful for destroying tumors . Further , gene therapies will rely on controlling the extent to which a therapeutic gene is delivered and expressed . Robust methods for controlling virus growth have yet to be discovered . However , for some viruses , such as vesicular stomatitis virus ( VSV ) , growth can be very sensitive to the specific linear order of its five genes . Our current work is significant in combining experiments and computational models to identify which virus genes and genome positions most sensitively impact VSV growth , providing a foundation for its applications in human health . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
]
| [
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"virology/viral",
"replication",
"and",
"gene",
"regulation",
"computational",
"biology/systems",
"biology"
]
| 2009 | Computational Fitness Landscape for All Gene-Order Permutations of an RNA Virus |
Understanding the rat neurochemical connectome is fundamental for exploring neuronal information processing . By using advanced data mining , supervised machine learning , and network analysis , this study integrates over 5 decades of neuroanatomical investigations into a multiscale , multilayer neurochemical connectome of the rat brain . This neurochemical connectivity database ( ChemNetDB ) is supported by comprehensive systematically-determined receptor distribution maps . The rat connectome has an onion-type structural organization and shares a number of structural features with mesoscale connectomes of mouse and macaque . Furthermore , we demonstrate that extremal values of graph theoretical measures ( e . g . , degree and betweenness ) are associated with evolutionary-conserved deep brain structures such as amygdala , bed nucleus of the stria terminalis , dorsal raphe , and lateral hypothalamus , which regulate primitive , yet fundamental functions , such as circadian rhythms , reward , aggression , anxiety , and fear . The ChemNetDB is a freely available resource for systems analysis of motor , sensory , emotional , and cognitive information processing .
Automatic keyword-based search ( 1 , 750 keywords ) and manual grey search on electronic databases revealed 124 , 694 abstracts , titles , or both identified as original publications . Out of these , 4 , 517 studies were relevant for data mining and data from 1 , 560 original research articles with 36 , 464 rats were selected for the connectome identification based on the inclusion criteria . A flow diagram of the study selection process is represented in supplementary information ( S1 Fig ) . According to our stringent inclusion and exclusion criteria , these selected studies represent the relevant published outcome of 55 years of neuroanatomical research . All selected studies were performed in outbred rats with no specific genotype or phenotype . Furthermore , animals did not receive any pharmacological treatment or behavioral training . On average , 86 . 08% ± 0 . 02% of the cases were male animals . The strain of the animals was inverse Gaussian distributed ( μ = 0 . 25;λ = 1 . 24 ) . Thereby , 54 . 03% ± 0 . 01% and 24 . 41% ± 0 . 01% of the animals were Sprague-Dawley and Wistar rats , respectively . In total , the systematic literature search identified 281 interconnected nuclei ( 4 , 585 axonal projections ) of cerebral regions , cerebellum , and components of medulla oblongata . The connectivity matrices associated with this raw database are presented in the supplementary information ( S7 , S8 and S9 Tables ) along with Bregma levels from rostral to caudal . The nomenclature-consistent , brain-wide ( cerebral ) core of the chemical connectivity database ( ChemNetDB ) comprises by 188 cortical and subcortical morphologically distinct regions with 3 , 712 connections . The database references are provided within S1 Text . This database considerably differs from the existing macroscopic connectomes [17 , 18] of the rat central nervous system . In contrast to ChemNetDB , which is a multiscale connectome database , these databases are often restricted to a single spatial scale . Furthermore , with the exception of a few brain regions such as amygdala , the bed nucleus of stria terminalis ( BNST ) , and cerebral cortex , they are only sparsely connected , and do not provide edge-complete connectivity . Databases lacking the edge-completeness property include brain regions that are either not connected to any other region , or receive only afferents from other regions , or send only efferents to other brain areas . None of these cases are biologically or physically feasible within the central nervous system . Thus , these databases effectively present cerebral subnetworks of the full connectome . In addition , concerns regarding nomenclature and data integration ( such as combining data from neonatal and adult animals ) and other inconsistencies make these platforms much more difficult to use . ChemNetDB overcomes these issues and shows significant advantages over the existing databases in that ( 1 ) it is currently the most comprehensive multiscale database that contains previous databases as subsystems and it is nearly edge-complete; ( 2 ) it integrates neurochemical information in a consistent and validated manner; and ( 3 ) it is consistent with respect to age of the animals and nomenclature ( see S2 Text for more detail ) . Divisive hierarchical algorithms with edge-completeness constraints further reduced the dimension of the network and integrated the extracted cerebral core into a multilayer cerebral neurochemical connectome of the rat brain . Neurochemical connections within and between brain regions were mapped into a 3-dimensional space using a standardized platform to generate a comprehensive and quantitative database of inter-areal and cell-type−specific projections . The connectome consists of 125 scale-consistent cerebral nuclei and 2 , 931 multiscale , multichemical connections ( Fig 1a ) . Thereby , 632 bidirectional , 1 , 642 unidirectional connections , and 25 loops utilize 25 different neurochemical compounds , such as amino acids , monoamines , cannabinoids , - opioids , and several other neurotransmitters . Despite such diversity , the chemical coverage of the network is only 25 . 08% and the majority of the connections were treated as binary links . Gamma-aminobutyric acid-ergic ( GABAergic ) neuronal connections ( Fig 1b ) constitute the dominant chemical components at the short-range ( 18 . 36% of the chemically denoted connections ) . In contrast , dopamine ( 15 . 37%; Fig 1c ) , serotonin ( 13 . 47%; Fig 1d ) , glutamate ( 10 . 75%; Fig 1e ) , and enkephalin ( 7 . 76%; Fig 1f ) represent the majority of chemical colocalizations of long-range axonal projections . A comprehensive systematic review on receptor distribution according to their mRNA expression levels and receptor binding was used to improve the robustness of these observations by providing spatial colocalization patterns of the transmitter systems within the connectome ( S10–S13 Tables ) . For this purpose data on glutamatergic ( mGluR1-7 , GluK4-5 , GluA1-4 , GluN1-2 ( A-D ) ) , monoaminergic ( D1-D5 , 5-HT1-7 ( A-C ) , β1–2 , α1-2 ( A-D ) ) , cholinergic ( M1-M5 , nAChRα ( 2–6 ) β ( 2–3 ) ) , GABAergic ( GABAAα ( 1–6 ) β ( 2–3 ) γ ( 1–3 ) δ , GABAB1 ( a , b , p ) , GABAB2 ) , opioid ( μ , δ , κ , ORL1 ) , and cannabinoid receptors from 103 ( out of 1 , 843 ) selected original research articles are qualitatively integrated into brain-wide receptor distribution density maps ( S2–S12 Figs ) . In the following , we provide examples of network analyses of the ChemNetDB . First , we investigated the network properties of the whole-brain connectome at 2 levels of resolution , namely as a coarser 19-node network ( see Methods ) described by the G19×19 adjacency matrix [12] and then at full resolution , as a 125-node network ( G125×125 ) . The G19×19 has a total of 236 directed links and thus a graph density of ρ19=23619×18=0 . 69 , similar to the interareal networks in the macaque and the mouse [19 , 20] . The 125-node network G125×125 has 2 , 906 directed links and a graph density of ρ125=2906125×124=0 . 19 , significantly lower than the coarse version . Both graphs are relatively small and the degree distributions are noisy ( Fig 2a and 2b ) , hence it is difficult to identify these distributions by simple fitting . However , below we present a model that generates predictions for these distributions ( black lines in Fig 2a and 2b ) . Plotting the degree sequence in Fig 2c and 2d ( rank ordering the nodes by their in- and out-degrees ) , we observe the existence of a small number of high degree nodes ( hubs ) receiving and/or sending many connections to the rest of the network . The BNST comprises 21 distinct brain areas , which are responsible for integration of limbic information and valence monitoring , processing threat reaction , fear , anxiety , and many other functions , collect 178 in-links , and project 365 out-links to the rest of the connectome . The dorsal raphe nucleus ( DRN ) has the largest number of outgoing projections ( 83 ) and the fourth largest of incoming links ( 53 ) . This correlates with the fact that the DRN is the largest provider of serotonin innervation to the rest of the brain . Similarly , infra- and prelimbic cortices that comprise prefrontal cortex have high out-going projections ( each 61 ) and a relatively high total degree of 105 . This corresponds with the key role of prefrontal cortex in regulating cognitive functions . Moreover , another hub is the lateral hypothalamic area ( with 52 in-degree and 60 out-degree ) , which is responsible for a significant array of functions , such as feeding behavior , wakefulness , thermoregulation , gastrointestinal functions , energy homeostasis , and visceral functions . Supplementary tables show the top list of degrees ( S1 and S2 Tables ) and betweenness centralities ( b . c . s; see Methods ) for nodes ( S3 and S5 Tables ) and edges ( S4 and S6 Tables ) at both resolution levels , and Fig 2e–2h show the distributions of the betweenness values . According to these lists , DRN has the largest node betweenness , and that , expectedly , correlates with its high in- and out- degrees , a property that holds in general , although with some exceptions , such as the central nucleus of the amygdala ( CeA ) . CeA , with the second highest b . c . , has the largest in-degree , but it is only 19th in the out-degree list . The fact that it plays a key role in information transmission and processing resonates with the observation that it is a principal area for controlling emotional reactions . The distributions of both node ( Fig 2c and 2d ) and edge betweenness ( Fig 2e–2h ) show a concentration at large values , suggesting a heterogeneous network structure; in particular , there are 8 edges with exceptionally high b . c . Given that for both macaque and mouse the interareal network is heterogeneous and shows a strong core-periphery structure [19–21] , we have analyzed whether the same holds in the rat brain as well . By using a stochastic block modeling [22 , 23] method ( see Methods section ) , we determined the probabilities of the individual nodes to belong either to the core or periphery ( Fig 3a–3c ) . In Fig 3 the core nodes are colored red while the periphery nodes are blue . It shows a clear separation of the two classes in a similar fashion to the macaque and mouse . For G19×19 the core contains 12 nodes with an internal density of ρcc = 0 . 87 , a periphery of 7 nodes of internal density ρpp = 0 . 24 , and the subgraph of edges running between core and periphery nodes of density ρcp = 0 . 66 . For G125×125 the core has 69 nodes with a high internal density of ρcc = 0 . 41 , a periphery of 56 nodes of internal density ρpp = 0 . 03 and the subgraph of edges running between core and periphery nodes of density ρcp = 0 . 12 . The list of core areas and the corresponding regions ( in the 19-node segmentation ) they belong to is shown in S13 Fig . Fig 3e shows the location of core nodes within G125×125 , using a force-based layout representation . Note that while this analysis clearly indicates the core-periphery organization of the connectome , it does not reveal its internal structure , in particular , the internal network communities , if any . To explore this aspect , we performed a hierarchical decomposition of the network using the Girvan-Newman algorithm ( see Methods ) , which outputs a hierarchical community dendrogram [24] . The results are shown in Fig 3b and 3d for both resolution levels . They indicate that in contrast with social networks , which have a clustering of communities over several levels , the rat connectome ( at either resolution ) has an onion-type structural organization [25] , in which layers of areas are added on top of the previous ones . On the x axis , we colored the areas according to their core-periphery membership , which correlates well with depth in the dendrogram , with core nodes concentrating towards the inside of the onion structure . While this overall onion structure is a direct consequence of high-network density , the membership of the levels from the innermost to the outer layers is highly specific to the network . Although the densities of G125×125 and its core are not drastically different ( 0 . 19 versus 0 . 41 ) , at this resolution such difference implies significant specificity . Specificity can be estimated by computing an upper bound to the probability that an Erdos-Renyi random graph on 125 nodes of density p = 0 . 19 will have a core on 69 nodes of internal density ρcc = 0 . 41 . The computation is shown in the Methods section , giving a very low probability of approximately 10−228 , showing that the rat connectome core-periphery structure is , indeed , highly specific . The raphe nuclei , DRN and medial raphe nucleus ( MRN ) , constitute the deepest core of the dendrogram or the heart of the onion structure . These serotonergic systems play an important and generalized role in regulation of sleep-wake states and behavioral arousal . While MRN is mostly involved in stress- and anxiety-related processes [26] , the DRN is critically involved in the neuronal regulation of circadian rhythms and sleep [27] , which may support the hypothesis that DRN acts as a pacemaker of the network . A prerequisite of controlling a complex biological system is its structural controllability [28–30] . As another example analysis , here we identify the minimal set of driver nodes that could be used to control the system , through searching for a maximum matching in the graph ( see definitions in Methods section ) . A realization of maximum matching is shown in Fig 4a with four driver nodes: subthalamic nucleus ( STh ) , subiculum , CA1 , and medial nucleus of amygdala . The interesting finding here is the role of STh in the structural controllability of the rat brain . Numerous deep brain stimulation studies have already shown the remarkable effects of activation of neurons within STh on global circuit dynamics [31] and in the treatment of Parkinson disease and other disorders [32] . This observation suggests an agreement between structural and functional controllability of neuronal networks and attracts our attention towards hippocampal and amygdaloid regions as potential targets for deep brain and other stimulation techniques . While the number of driver nodes ( 4 ) is small compared to the total number of 125 nodes of the network , it is still significantly larger than in random networks with similar density . To show this we performed 500 randomizations of the network preserving the degree sequence and obtained an average of 1 . 23 driver nodes ( SD: 0 . 45 ) . Sequentially removing the longest ( weakest ) edge from the network , the number of driver nodes increases , as expected , exponentially ( Fig 4b ) . However , the number of driver nodes for the original network , is consistently larger than for its degree preserved randomized versions . Thus , the original network structure is such as to allow for more points of control in the network , that is for a larger control diversity , than in a similar random graph with the same degree sequence . Consistent , cortex-wide retrograde tracer injections in several species ( macaque [20 , 33] ) , mouse and microcebus [19] have shown that the distribution of the lengths of white matter ( WM ) axons follow an exponential decay , called exponential distance rule ( EDR ) , with a species-dependent decay rate ( λ ) . Naturally , 2 questions arise: Is the rat large-scale connectome also described well by the EDR network model ? And ( 2 ) Does the agreement with the EDR model break with increasing resolution ? The second question could not be answered for the other species , as there is no ( nearly ) edge-complete database available at higher resolutions for them , and thus ChemNetDB is especially valuable in this regard . By using ChemNetDB , we next attempt to answer both questions by modeling the connectome with the EDR network model at the G19×19 and G125×125 levels , respectively . Network comparison is based on parametric property matching described briefly in the Methods section and in detail in refs [19 , 20] . Fig 5 shows the results for the G19×19 using a set of commonly used graph measures: number of uni- and bidirectional edges ( Fig 5a ) , the root-mean-square ( RMS ) of deviations for the 3-motif counts between model and data ( Fig 5b , 5f and 5g ) , the RMS of clique-count deviations between model and data ( Fig 5c and 5h ) , the RMS of the deviation of the eigenvalues of the co-occurrence matrix AAT between model and data networks ( Fig 5d and 5i ) and finally the clustering coefficients ( Fig 5e ) . These results show that all models generate the same value for the decay rate λ in the range 0 . 60 − 0 . 65mm−1 consistently , indicating that the EDR network model is a good model for the rat connectome at the large-scale level . Fits generated by and EDR with λ = 0 . 6mm−1 are shown in Fig 2a and 2c for the degree distributions and the degree sequence , respectively . Moving to the higher resolution connectome G125×125 , the same procedure yields the comparisons shown in Fig 6 . The figures now show a different picture: the best λ values determined from parameter matching are varying , from measure to measure , indicating that a single λ-parameter EDR model cannot describe the whole data network at this resolution . The reason lies with the fact that once we subdivide brain regions ( as it was done in going from 19 areas to 125 ) , there will be an increasing number of area pairs that are connected by gray matter ( GM; nonmyelinated ) connections , instead of WM connections . Note that at the G19×19 level , the connections are WM connections , just as for the cortical interareal networks in the mouse and macaque , where the EDR descriptions work well . As shown by experiments presented in Horvát et al . [19] , the decay rate is sensitive to the nature of the medium in which the connections are running ( WM versus GM ) . In particular , while local ( GM ) connections obey an EDR with almost the same decay rate ( 4 . 6 ÷ 4 . 9mm−1 ) in macaque , mouse and rat ( see Fig 11B in [19] ) , for WM connections , the EDR decay rate decreases with increasing brain size ( 0 . 19mm−1 for macaque , 0 . 8mm−1 for mouse , and 0 . 6mm−1 for rat—this paper ) . The larger the brain , the smaller the decay rate for nonlocal ( WM ) connections . The EDR network model is defined via a single decay rate and it works well for all those brain networks for which the connections obey a single parameter EDR . Once the network contains both types of connections ( GM and WM ) , as the rat brain at 125-area resolution , a single decay rate model will not work as shown in Fig 6 . To develop a two-parameter EDR model , however , we also need to include information about the location of the GM connections with respect to the WM ones , resulting in a more involved model , which will be the subject of a forthcoming paper .
Most studies in neurobiology rely on a precise understanding of the neuronal connectivity and its neurochemical actors . Yet , investigators commonly face massive data that require enormous resources to be processed for their demands . By utilizing advanced neuroinformatics , our study resolves this problem and integrates over 50 years of neuroanatomy research on rat brains into a consistent multiscale , multilayer neurochemical cerebral connectome . Supervised machine learning was applied to resolve nomenclature issues resulting in an extensive standardized database , which combines the state-of-the-art knowledge of connectomics and neurochemistry . Establishing the neurochemical connectivity database ( ChemNetDB ) is a novel approach towards topological mapping of the brain that takes the neuroconnectomics well beyond the binary constructs and paves the way for more advanced and accurate investigations of healthy and disease states of rat brains . We have investigated , as examples of analytic studies that could be done on the database , several network measures and their relationships with similar analyses in macaque and the mouse . Analysis of the structural properties of the resulting network reflect and confirm the key functional roles of deep subcortical brain areas such as lateral hypothalamus , BNST , and DRN . These regions are known to be responsible for primitive yet essential and evolutionary conserved functions , such as regulation of sleep and circadian rhythms , reward , anxiety , aggression , and fear . The importance of these brain areas for survival is also associated with an early formation in the developmental stages as reflected in the connectome by their extremal values of various graph connectivity measures such as b . c . and/or node degree . Previous theoretical analysis demonstrated that the optimal core-periphery structures in networks that are stable against both random and targeted failures/attacks ( targeting hubs ) are onion-like structures [34] , however , they have not been observed in real-world networks [25] . Our analysis shows that the onion structure occurs in large-scale , dense brain neuronal networks , where robustness against information transmission failures ( of all types ) is a critical requirement . The observed onion structure does not contradict the well-known hierarchically modular structure of brain networks . At this large-scale resolution , applying the Girvan-Newman methodology [24] , the paths-based network modularity is not revealed due to the high density of the network . Instead , the onion organization is dominant and the community structure appears as core-periphery at this scale . As another example of an analytical study on ChemNetDB , we tested whether the EDR property and the associated network model are consistent with the rat connectome structure , at different levels of resolution . The EDR , expressing economy of wiring has been found to hold in both macaque and mouse for both mesoscale WM connections and local GM connections . While direct measurements of the WM axon length distribution in the rat are currently lacking , the EDR network is highly consistent with the large-scale ChemNetDB connectome , suggesting that the whole-brain WM connectivity is also strongly determined by the EDR . By using maximum matching algorithms , we have identified 4 nontrivial driver nodes of the network . In particular , STh appears as a feasible candidate to be involved in control mechanisms of the brain . However , the actual biological interpretation of driver nodes is not yet clear and the role of the identified brain regions in controlling brain activity requires further investigations and experimental validations . Note that the outcomes of the example network analyses presented here might change with the addition of new data; however , the main observations , such as the strong core-periphery structure and network measures expressed as fractions ( including density , the fraction of 3-motifs , etc . ) , are expected to be robust against changes due to the high density of the networks . ChemNetDB provides the first , whole-brain , large-scale and consistently collated rat connectome database that also includes neurochemical specificity . This will provide researchers with a tool to gain insights into the fundamental relationships between connectome architecture , information processing , and brain function , with potential for advancing preclinical research and clinical applications such as those related to substance abuse and depression . | The mammalian brain consists of a network of chemically diverse , multiscale , and multilayer neuronal wiring patterns that form the physical infrastructure underlying the processing of motoric , sensory , emotional , and cognitive information . Decades of histological track-tracing studies have generated a set of highly valuable yet disorganized data , which is very hard to retrieve in a hypothesis-free manner . We present an open-access database ( ChemNetDB ) that organizes over 50 years of neuroanatomical track-tracing and neurochemical measurements from 36 , 464 rats . This neurochemical connectome is terminologically consistent and shares several network invariants with mouse and macaque cortical networks , suggesting that the mammalian brain exhibits universal structural features . Furthermore , several network measures reflect and confirm the key functional roles of deep subcortical brain areas , which are known to be responsible for primitive , yet essential and evolutionary conserved functions related to survival . ChemNetDB provides the first whole-brain , multiscale , and consistently collated rat connectome database . ChemNetDB also includes neurochemical specificity and represents a powerful tool for in vitro , in vivo , and in silico investigations of brain function and disorders . | [
"Abstract",
"Results",
"Discussion"
]
| [
"medicine",
"and",
"health",
"sciences",
"methods",
"and",
"resources",
"neural",
"networks",
"nervous",
"system",
"vertebrates",
"neuroscience",
"animals",
"mammals",
"data",
"mining",
"primates",
"animal",
"models",
"model",
"organisms",
"mathematics",
"brain",
"mapping",
"network",
"analysis",
"experimental",
"organism",
"systems",
"information",
"technology",
"directed",
"graphs",
"old",
"world",
"monkeys",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"monkeys",
"animal",
"cells",
"graph",
"theory",
"macaque",
"connectomics",
"rodents",
"cellular",
"neuroscience",
"neuroanatomy",
"anatomy",
"cell",
"biology",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"amniotes",
"organisms",
"rats"
]
| 2017 | A multiscale cerebral neurochemical connectome of the rat brain |
Changes in synaptic efficacy are believed to form the cellular basis for memory . Protein synthesis in dendrites is needed to consolidate long-term synaptic changes . Many signals converge to regulate dendritic protein synthesis , including synaptic and cellular activity , and growth factors . The coordination of these multiple inputs is especially intriguing because the synthetic and control pathways themselves are among the synthesized proteins . We have modeled this system to study its molecular logic and to understand how runaway feedback is avoided . We show that growth factors such as brain-derived neurotrophic factor ( BDNF ) gate activity-triggered protein synthesis via mammalian target of rapamycin ( mTOR ) . We also show that bistability is unlikely to arise from the major protein synthesis pathways in our model , even though these include several positive feedback loops . We propose that these gating and stability properties may serve to suppress runaway activation of the pathway , while preserving the key role of responsiveness to multiple sources of input .
Protein synthesis is a necessary stage in long-term synaptic plasticity , both for strengthening synapses ( long-term potentiation , LTP ) and weakening them ( long-term depression , LTD ) [1] , [2] . A substantial part of this synthesis occurs in the dendrites , and possibly close to the modified synapse ( s ) [3] . mRNAs are transported to the dendrites where local sites of protein synthesis machinery appear to serve a small number of synapses or dendritic spines [4] . There are now several putative signaling pathways that connect synaptic input to protein synthesis [5] , [6] . The relationship between synthesis triggered by activity , different receptors , and growth factors is important in defining the logic of memory formation [7] , [8] . Dendritic protein synthesis is a particularly obvious example of a positive feedback process , since many of the newly synthesized proteins either provide signal input to the system , or are part of the synthesis machinery itself . From the viewpoint of memory formation , positive feedback is a significant feature: it suggests that self-sustaining activation processes may occur [9] . Such self-sustaining processes frequently take the form of bistable switches , and indeed a protein synthesis switch has been proposed to sustain memory in Aplysia [10] . Many biochemical and signaling switches have been proposed at the mammalian synapse , for example , calcium calmodulin type II kinase ( CaMKII ) autophosphorylation [11] , MAPK-PKC feedback [12] , [13] , and receptor cycling [14] . Given its known role in synaptic plasticity , it is tempting to consider the possibility that protein synthesis may also form such a switch . Synaptic protein synthesis has attracted considerable recent attention , and several of its key pathways have been identified . Major inputs include neurotrophins such as BDNF [15] and neurotransmitters such as glutamate , which in turn may act through metabotropic [16] and ionotropic [17] pathways . The BDNF pathway has been extensively studied and at least one rather lengthy signaling cascade has been identified that culminates with protein synthesis regulation ( Figure 1 ) . In this cascade , BDNF binds to its receptor , tropomyosin-related kinase B ( TrkB ) , which activates phosphatidylinositol-3-kinase ( PI3K ) , then AKT , which elevates Rheb-GTP to activate the mammalian target of rapamycin ( TOR ) . TOR acts on synthesis both through its phosphorylation of the S6 Kinase ( S6K ) , and through its effects on eukaryotic initiation factor 4E ( eIF4E ) . Both of these are involved in translation initiation and hence control protein synthesis . The metabotropic glutamate receptor ( mGluR ) regulation of synthesis is less understood , but there is evidence that it joins this same cascade at AKT [18] , though the upstream steps are yet to be resolved . Ionotropic-glutamate signaling works through the association of postsynaptic depolarization and presynaptic glutamate , leading to opening of the N-methyl-d-aspartate receptor ( NMDAR ) and calcium ( Ca2+ ) influx . Calcium has multiple effects on protein synthesis . Two of the possible inputs are via Mitogen-activated protein kinase ( MAPK , present in neurons as Extracellular signal-regulated kinase II , ERKII ) and via calcium-calmodulin type III kinase ( CaMKIII ) . MAPK , like mTOR acts on S6Kinase and 4E-binding protein ( 4EBP ) which blocks eIF4E . CaMKIII phosphorylates eukaryotic elongation factor-2 ( eEF2 ) and inhibiting its elongation activity ( Figure 1 ) . These downstream targets ( eIF4E and 40S ) represent possible control points for selective protein synthesis . This is because 40S phosphorylation has been particularly implicated in the translation of a subset of mRNAs with a 5′ terminal oligopyrimidine sequence ( 5′TOP ) . 5′TOP proteins are involved in protein synthesis , and the 40S ( S6 , a subunit of 40S ) protein itself has a 5′TOP [5] . In the absence of such 5′TOP targeting , other dendritic mRNAs are translated , including Arc , αCaMKII , microtubule-associated protein 2 ( MAP2 ) , neurogranin [5] and interestingly BDNF itself [19] . In this study we explore this convergence of three inputs: BDNF , MAPK , and Ca2+ onto dendritic protein synthesis . We first parameterize the complex BDNF signaling pathway in stages , using published experimental data . We fold in previously published models of Ca2+ and MAPK signaling , and check that the combined model is consistent with further experiments . With the combined model we explore a range of input combinations and analyze how the positive feedback in the system may influence its behavior .
We incorporated previously published models for CaM , PKC and MAPK ( ERKII ) signaling inputs to the current model [55] ( Figure S2 ) . These models have all been parameterized and matched with experimental data in our earlier work . The main input to these pathways in our model is Ca2+ , which stimulates PKC directly . MAPK is downstream of PKC as well as CaM-Ca4 in our model , so MAPK activity is closely tied to synaptic activity leading to Ca2+ influx . Many additional inputs to MAPK are known but were not included in our model . We merged all the sub-models mentioned above and tested our overall model . In contrast to the individual constraint steps for each sub-model , the composite model validation monitored signal flow through the entire cascade . We illustrate the pathway with a block diagram ( Figure 5A ) , where the readout points are in gray . We first modeled a time-course experiment in primary cultures of rat cortical neurons , in which a sustained 2 nM BDNF stimulus was applied in the bath , leading to activation of AKT ( Figure 5B ) [20] . The simulated time-course was slightly slower than the experimental time-course . We attributed this to the difference in temperatures between the constraining experiments used for our model ( ∼30 degrees C ) and the temperature of this experiment ( ∼37 degrees C ) . We then modeled the published experiments which measured the relative increase in several downstream readouts and protein concentration following steady BDNF stimulation at 3 . 7 nM ( 100 ng/ml ) [56] . Again , these experiments were carried out using primary cultures of rat cortical neurons at 37 degrees C . Therefore , we ran the simulations for twice as long ( TSC1-TSC2* , P-4EBP , eIF4E-4E-BP , P-S6K , P-S6 for 10 minutes and protein for 30 minutes ) as in the experiment ( i . e . for 5 minutes and 15 minutes respectively ) before measuring the readouts ( Figure 5C ) . Finally , we validated our model against experiments carried out in primary hippocampal cultures from control and MAPK knockout mice , again at 37 degrees C [57] . These experiments measured intermediates and protein synthesis responses to BDNF stimulation . Our model was able to closely match the experimental observations ( Figure 5D ) . Overall , our composite model incorporated several individually constrained pathways , and we showed that the combined model was in agreement with more complex experiments that exercised a significant part of the protein synthesis regulatory cascade . Our data sources were diverse , and despite our modular approach to parameter fitting , we had to estimate many rates indirectly . We therefore tested how parameter uncertainty affected model behavior by systematically varying each parameter and comparing responses with those of the original model . This approach reflects the robustness of the in vivo system where a small change in conditions does not lead to much change in its behavior , unless the change impinges on a key cellular control molecule . We systematically varied two enzyme parameters ( Michaelis constant ( Km ) ( Figure 6B ) and turnover number ( kcat ) ) ( Figure 6C ) , two reaction parameters ( forward rate ( Kf ) ( Figure 6D ) and backward rate ( Kb ) ( Figure 6E ) ) , and the total concentrations of each molecule ( CoInit ) ( Figure 6A ) . Each parameter was varied over 2 log units , from 0 . 1 to 10 times the reference model value . As a readout we monitored the steady-state levels of PIP3_AKT-t308_s473 ( active AKT ) , 4E-BP_tot ( sum total of the phosphorylated forms of 4E-BP ) , MAPK* , S6K_tot ( sum total of the phosphorylated forms of S6K ) ( Figures S3 , S4 , S5 , S6 , S7 ) and protein synthesis ( Figure 6 ) . Most parameters introduced only small changes in these readouts . A handful of parameters resulted in a two-fold or greater effect on responses , and these fell into two sets , corresponding to the BDNF and MAPK pathways of the model . For the BDNF pathway , as measured by AKT activation , the major control molecules were PI3K , PTEN , PIP2 and PP2A . Most enzyme and binding reactions that were sensitive were also associated with these molecules . For the MAPK pathway , the control molecules were GEF , Ras , GAP , craf-1 , PP2A and MKP-1 . In addition to the reactions involving these molecules , the Ca2+ entry and PKC activation steps were sensitive control parameters . A final key control parameter was the background rate of phosphorylation of S6K ( S6K* ) ( Figure 7C ) . This is a convergence point for many kinases , including MAPK , SAP kinases , p38s and Cdc2 [38] . Of these , only MAPK is included in the present study . As discussed below , this phosphorylation step may act to switch the protein synthesis pathway into two distinct modes of responses to inputs . In a first set of predictions of the composite model , we examined steady-state responses to BDNF and Ca2+ stimulation . The response to BDNF was a typical sigmoid but with a rather high baseline , so that the total increase in protein synthesis rate was just over two-fold ( Figure 7A ) . High Ca2+ lowered the response and nearly flattened the sigmoid ( Figure 7B ) . We then investigated why Ca2+ acted in this unexpected manner . From the reaction scheme , Ca2+ stimuli led to activation of two key kinases with opposing effects ( Figure 7C ) . The MAPK cascade was activated through CaM-Ca4 as well as PKC . BDNF had no effect on MAPK responses . CaMKIII was also activated by CaM-Ca4 , but at a higher Ca2+ concentration ( Figure 7D ) . This opposing effect suggests that there may be a narrow window of Ca2+ in which synthesis is maximal . This narrow window gives rise to a bell-shaped curve when protein synthesis is plotted against Ca2+ levels . Interestingly , the peak of this bell curve is at rather low calcium levels , only about 2 or 3 fold basal levels . Mechanistically , this is because the differential region of the two Ca2+ activation curves ( Figure 7C ) is in this Ca2+ range . Physiologically , this Ca2+ range corresponds to mild rather than intense synaptic input [58] . In our default model , we found that high Ca2+ above 0 . 5 µM turned off protein synthesis ( Figure 7E ) . As expected , higher levels of CaMKIII lowered this cutoff point ( Figure 7F ) . Although these responses might suggest that strong Ca2+ influx ( as in LTP stimulation ) should block synthesis , the CaMKIII response is much faster than that of MAPK , and the system response to a Ca2+ transient may be more nuanced . We test this situation in the section on LTP and LTD stimuli . One of the key regulatory input points to the protein synthesis cascade is S6K phosphorylation . S6K is known to be a target for other kinases , including SAP kinases , p38s and Cdc2 [38] , in addition to MAPK and TOR ( Figures 3A and 7C ) . In the model we had a steady basal kinase ( BK ) activity ( rate is equal to 0 . 01 /sec ) ( Figure 7F ) to represent the external kinase inputs . We now varied this BK activity to represent regulatory input . We found that the BK rate had a profound effect on the network controlling protein synthesis . At low BK , the network was in a state where protein synthesis was mostly low and gradually increased with an increase in MAPK activity . At high BK , the network was less susceptible to change in MAPK activity ( Figure S8 ) . Second , Ca2+ sensitivity switched from an inhibitory response above 0 . 5 µM Ca2+ to a BDNF-gated bell-shaped response ( Figure 7G and 7H ) . Thus protein synthesis was high only when BDNF was high , and Ca2+ was in the 0 . 15 to 0 . 3 µM range . This unusual profile was due to high-sensitivity activation of MAPK by Ca2+ , followed by the lower-sensitivity inhibition of protein synthesis via CaMKIII ( Figure 7D ) . Overall , we found that our composite model exhibited two possible kinds of behavior: BDNF gated and MAPK gated . The BDNF-gated behavior seemed more consistent with experiments in neuronal systems , and was contingent upon background phosphorylation of S6K by kinase regulation outside the scope of our study . In this mode BDNF had a strong effect on protein synthesis , and MAPK ( stimulated by Ca2+ in our model ) acted to further elevate responses . The MAPK-gated behavior may be of interest in other cell types [59] , [60] or specific neuro-regulatory contexts [61] . It had a lower basal synthesis rate , a strong dependence on MAPK , and a narrow window of activation by Ca2+ . We next tested the model responses to temporal input activity patterns used for inducing LTP and LTD . We combined patterned Ca2+ input with varying levels of simulated BDNF to explore the effects of possible combinations of these inputs at the synapse . Based on the pathways in our model , BDNF should activate synthesis through AKT and TOR , whereas Ca2+ should turn on MAPK to activate synthesis , but also turn on CaMKIII to depress synthesis . We represented the LTP stimulus as three pulses of Ca2+ influx , each 1 second wide , and separated by 300 sec [55] . In addition , we provided a BDNF input of 3 . 7 nM for 5 sec for each pulse of Ca2+ ( Figure 8A ) . We monitored MAPK* , AKT* ( PIP3_AKT_thr-308 ) , CaMKIII* , and protein synthesis rate levels . The LTP stimulus caused a modest and brief elevation of protein synthesis rate through the combined action of MAPK and BDNF ( Figure 8A , 8C , 8E , and 8G ) . We compared responses with three pulses of 10 µM Ca2+ , and at basal Ca2+ ( 0 . 08 µM ) . We found that the MAPK response was entirely Ca2+ dependent . At basal Ca2+ CaMKIII and protein responses were indeed present , and were about half as large as for the 10 µM Ca2+ stimulus . This suggested that the contributions of the BDNF and Ca2+/MAPK inputs to protein synthesis were about equal . We modeled LTD input as a single 900 sec Ca2+ pulse [62] along with a BDNF elevation to 3 . 7 nM for the same period . We again monitored MAPK* , AKT* ( PIP3_AKT_thr-308 ) , CaMKIII* and protein synthesis rate ( Figure 8B , 8D , 8F , and 8H ) . Interestingly , the MAPK response was much stronger for LTD than for LTP stimulus , and the protein synthesis levels were nearly twice as large . BDNF had exhibited gating behavior for steady-state responses , so we tested if it also did so for these transient stimuli . We ran additional simulations with zero BDNF , and found that though there were transients above and below baseline; neither the LTP nor LTD Ca2+ stimulus resulted in any net synthesis ( Figure S9 ) . Thus BDNF ‘gates’ LTP/LTD stimulated protein synthesis . While we did not model mGluR explicitly , it is known to converge onto AKT [18] which is in the BDNF pathway as well , and upstream of TOR in our model . Activation of TOR strongly enhanced the ability of both LTP and LTD stimuli to cause protein synthesis . Thus mGluR input might also act as a gate . In contrast to the steady-state results , the large but brief Ca2+ transients in HFS stimuli caused an increase in protein synthesis provided BDNF was present . Our simulations showed that the transient stimulus did cause a brief CaMKIII-mediated reduction in synthesis , but this inhibition was relieved as soon as Ca2+ returned to baseline ( Figure 8E and 8F ) . The same Ca2+ transient also acted through the slower MAPK-eIF2E pathway , leading to an elevation of synthesis after the inhibition had subsided . Thus the protein synthesis cascade acted like a transient detector for Ca2+ inputs . There are also CaMKII-mediated inputs to protein synthesis which may provide further timing interactions , but we were unable to parameterize these mechanisms on the basis of current data [58] . The duration of protein synthesis for LTP and LTD stimuli was short: of the order of 30 minutes . This is much less than the time-scale of protein synthesis dependent plasticity responses . We consider this interesting discrepancy in the discussion . How might LTP and LTD stimuli differentially affect protein synthesis ? One possibility is the differential phosphorylation of S6 leading to the activation of the 40S subunit , which has been implicated in 5′TOP mRNA translation [5] , [63] . Another selective pathway is the eIF4E phosphorylation , which is elevated for CAP protein production [64] , [65] . It is believed that MAPK phosphorylates MAPK-interacting kinase ( MNK ) which in turn phosphorylates eIF4E [5] , but our model did not include MNK . There was no direct estimate for eIF4E phosphorylation so we took active MAPK ( which phosphorylates eIF4E ) as readout for CAP-protein production . We monitored peak levels of 40S and active MAPK complex ( as an estimate of CAP-protein production ) for LTP and LTD simulations , while varying the BDNF stimulation over the range 0 to 3 . 7 nM in each case . This exploration was necessary because we did not have direct data for BDNF release levels during different synaptic stimulus protocols . We found that the MAPK response was independent of BDNF ( Figure 8J ) , but the 40S levels were strongly BDNF dependent ( Figure 8I ) . Except for very low levels of BDNF , the LTD stimulus elicited a nearly 2-fold larger 40S response than the LTP stimulus . Overall , we find that TOR activation by BDNF is a prerequisite for both LTP- and LTD-triggered protein synthesis in our model . Different activity patterns , and different levels of co-regulatory inputs such as BDNF , may lead to a spectrum of differential activation of 5′TOP mRNA translation . There are at least two forms of feedback possible in this system . First , protein synthesis may increase the production of the protein synthesis machinery itself . The ribosomal 40S subunit and the eEF2 protein are examples of such feedback molecules in our model ( Figure 9A ) . Second , protein synthesis may increase the production of molecules , including BDNF , that contribute to activating synthesis ( Figure 9A ) . BDNF is released from neurons in an activity-dependent manner , and recent studies show that some of this release may occur through postsynaptic mechanisms [66] , [67] . Furthermore , there is evidence that BDNF has a postsynaptic site of action , in the induction of LTP [68] . Several studies have suggested that this release , coupled with the role of BDNF in synaptic plasticity , may constitute a feedback loop , and that at least part of this loop may be postsynaptic ( reviewed in [69] ) . While it is likely that there is a combination of pre- and postsynaptic BDNF feedback in its overall action , for the purposes of our analysis we have considered only the postsynaptic component . Both the protein synthesis machinery feedback , and the BDNF feedback have a positive sign , and are therefore suggestive of bistability [70] , [71] . While 40 to 400 species of synaptic molecules have been reported to be synthesized at the dendrite [72] , our model only included these three putative feedback molecules: BDNF , eEF2 , and 40S . We first considered synthesis control of each of these individually , then in combination . A simple graphical way to identify bistable systems arising from chemical feedback is to use overlaid dose-response curves [12] , [73] . Intersection points on such curves indicate stable points of the system . If there are three intersection points ( e . g . , Figure 9B between the sigmoid and middle straight line ) then the system is bistable . The lower intersection point is a stable state of low activity , and the upper intersection point a stable state of high activity . The intermediate intersection point has properties of a threshold or transition point between the two stable points . Note that these are steady-state curves , and the time taken to actually change state depends on reaction kinetics , the initial conditions , and on the kind of stimulus . In our study the concentration of the synthesized protein ( BDNF , eEF2 or 40S ) is directly proportional to our readout of protein synthesis rate , assuming that molecular degradation is first order in concentration . In the following analysis , we therefore used the protein synthesis rate on the y axis as a surrogate for concentration . This has the advantage that two unknowns can be folded together into a single scaling factor: the degradation rate and the fraction of the specific protein out of the total protein synthesized . In the case of protein synthesis feedback , the simplest assumption was that a fixed fraction F of the synthesized protein could feedback to increase synthesis . Thus , protein concentration [X] may be represented as a straight line of slope 1/F through the origin ( dashed lines in Figure 9B and 9C ) . It is clear from the geometries that a sigmoid dose-response curve may exhibit protein-synthesis bistability for some values of F ( Figure 9B ) , but a logarithmic or saturating curve will never be bistable if it has a linear dependence on synthesis rate ( dashed line in Figure 9C ) . The only situation in which a molecule with a saturating response curve might exhibit feedback bistability is if the level of protein is a steep sigmoid function of the synthesis rate ( smooth line in Figure 9C ) . We ran dose response curves for BDNF , the eEF2 protein and the ribosomal 40S subunit to look for bistability in this system ( Figure 9D–F ) . None of the curves could support bistability with the linear feedback assumption ( straight lines ) . In each case , there was just one intersection point between the dose-response curves , for any value of F . This was the case even if the system was co-stimulated with 3 . 7 nM BDNF ( Figure 9E and 9F ) . Having shown that individual synthetic pathways were unlikely to lead to bistability , we then asked if combinations might do so . We first considered higher-order activation processes , using the BDNF response as a typical dose-response curve . A hypothetical mechanism for this might be if the molecule acted in a higher-order manner to stimulate protein synthesis . We found that a fourth-order reaction of the form of the BDNF curve was sufficiently sigmoidal to just support bistability , but a third-order reaction was not ( Figure 9G ) . As a more biologically grounded test , we considered if the known dendritically synthesized molecules BDNF , eEF2 , and 40S might act synergistically to support bistability . We took an ‘optimum’ condition , where the protein synthesis scaling factor F for each was such that their half-maximal levels ( Khalf ) coincided . Using this scaling we re-ran the dose-response simulation , and found that the resultant curve was indeed sigmoidal and could support bistability for a narrow range of synthesis scaling factors ( Figure 9H ) . As we discuss below , this bistability requires a finely tuned set of conditions to occur , but on the other hand may be strengthened by other , as yet unknown feedback reactions . Overall , we conclude that a protein-synthesis switch at the dendrite is unlikely with currently known feedback mechanisms .
Bistable switches are plausible mediators of long-term storage of information at a signaling level . Several such switches have been proposed to be involved in synaptic plasticity: CaMKII autophosphorylation [11] , MAPK feedback loops [12] , [13] , [74] , and alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionate receptor ( AMPAR ) trafficking cycles [75] , Our expectation for this pathway was that it might exhibit switch-like behavior of its own , resulting from positive feedback from synthesized proteins back into the synthesis machinery . However , we found that bistability could not occur if we used the simple assumption that the each feedback molecule was present at a fixed fraction of the total protein synthesized . We were able to achieve a fragile bistability when we assumed coordinated feedback effects , but only with considerable tuning of feedback processes . At face value this is extremely unlikely situation . Furthermore , an elevation of basal synthesis rates , or influx of some of the 5′TOP proteins from non-local synthesis , would shift the synthesis curves upward along the y-axis , and thus eliminate the bistability . As a counterpoint , however , it is possible that synergistic effects may occur due to further feedback effects of additional dendritically-synthesized proteins . Overall , we feel that protein-synthesis-feedback bistability is unlikely , but it will take considerable additional data to resolve this point . While it seems reasonable that synaptic plasticity should draw upon new proteins and hence require elevated synthesis , recent studies suggest that the relationship between synthesis and plasticity is complex . Even though new synthesis is needed for plasticity , the total level of protein synthesis does not appear to change much following L-LTP stimuli [2] . Furthermore , background activity may actually act to suppress synthesis [76] . These observations further complicate the idea of protein-synthesis bistability in the dendrite . Our model of dendritic protein synthesis involves some 130 molecular species , including phosphorylation states and complexes of the key proteins . We have chosen this size model as a compromise between completeness and parameter unknowns . These concerns are present in all but the most exhaustively characterized signaling pathways [77] , [78] . We have selected a subset of implicated pathways for this study based on their reported roles , and on the degree to which the pathways have been characterized . For example , we were not able to include the mGluR input to the pathway as it is still lacking key functional details [18] . Nevertheless , our integrated model validation results suggest that we have succeeded in capturing many of the key interactions for the processes we wished to study . The resulting model was rather insensitive to parameter variations . Even the ‘sensitive’ parameters , when scaled up or down tenfold , affected model responses by barely a factor of two . Key targets for future refinement of the model , besides mGluR input , would be a more complete analysis of differential protein synthesis , and incorporation of the kinases that phosphorylate S6K on its p70 carboxy-terminal tail [38] . A much more challenging goal is to model the turnover of every protein in the model , so that all of the molecular concentrations are the outcome of regulatory and homeostatic controls . This awaits substantial advances in our understanding of all stages of neuronal protein synthesis control . The synapse is an unusual computational entity: it modifies its own hardware . Most computer languages manipulate data , but scrupulously avoid modifying program instructions . By these engineering standards , the synapse appears peculiar and unstable . In this system molecular signals ( data ) compute through chemical interactions ( program instructions ) , which among other things also structurally modify the synapse and thereby change its computational properties . The protein synthesis cascade is an instance of this self-modification . How does the cell avoid runaway processes ? Our study has some possible hints about how the system activity remains bounded . First , the synthetic increase is quite limited and the only way to get large fold activation over baseline is to provide synergistic BDNF and activity input . Thus , gating restricts the conditions where self-modification may occur . Second , our model suggests that the molecular logic of synthesis control may be somewhat insulated from the potentially dangerous process of positive feedback . Instead , potential feedback molecules seem to have saturating response curves with a high baseline that are effective at damping runaway buildup . With this interpretation , activity triggers leading to plasticity may initially activate biochemical switches rather than a protein synthesis loop . Candidate biochemical switches include CaMKII for fast switching , and MAPK or trafficking switches for slower phases of plasticity [12] , [75] , [79] . These switches , in turn , may regulate the protein synthesis pathway . Thus the short time-course of responses of the current model to synaptic input may simply reflect the role of synthesis as an effector of an upstream switch , rather than a switch itself . We speculate that very long-term changes may involve a shift in what is synthesized , and where it is made , rather than in how much synthesis occurs . To use our computational metaphor , this would be more like the machine loading different programs , rather than rebuilding and redesigning itself . This view would predict that very long-term plasticity involves decisions between a set of possible ‘programs’ executed by dendritic synthesis machinery to influence local synaptic function .
Simulations were developed in GENESIS , the General Neuronal Simulation System , using the Kinetikit interface [80] and solved using the exponential Euler method . Later parameter sensitivity , dose-response , and time-course runs were done using MOOSE , the Multiscale Object-Oriented Simulation Environment ( http://moose . sourceforge . net/ ) and solved using an adaptive Runge-Kutta method ( GNU scientific library ) . Simulations were carried out on PC workstations and on a SUN/Opteron cluster running Linux . Complete model reaction schemes and parameters are presented in Dataset S2 . To further check the model calculations , and to facilitate community access to our simulations , we converted our model to SBML Level 2 Version 1 ( Dataset S3 ) . We had to remove a tabulated BDNF-dependent calcium stimulus which is not supported by most simulators . We compared this slightly modified model with other simulators ( e . g , COPASI 4 . 4 [81] and CellDesigner 4 . 0 [82] ) and obtained the same results as with GENESIS and MOOSE . To validate our composite model , we ran the simulation for 6000 sec without any stimulus and then noted the concentration of TSC1-TSC2* , 4E-BP_tot , S6K_tot and total synthesized protein . We then applied a steady stimulus of 3 . 7 nM BDNF . In the experiments , the levels of TSC1-TSC2* , 4E-BP_tot and S6K_tot were noted at 5 min and the level of protein at 15 min . As our simulations used room-temperature rates , and these experiments were done at physiological temperatures , we doubled the runtimes for our simulations when carrying out these validations . The activated values were divided by the basal values to obtain fold activation . Sensitivity analysis was done by scaling each parameter one at a time in the range of 0 . 1 to 10-fold of the original parameter values . The parameters were: initial concentration ( for molecules with non-zero initial concentrations ) ; Km and kcat ( for enzymes ) ; and Kf and Kb ( for binding/conversion reactions ) . To measure sensitivity , we ran the scaled model for 6000s without any stimulus . At 6000s the value of BDNF was set at 3 . 7 nM and then the concentration of readouts was recorded at 6600 sec and 9600 sec . The concentration of the readouts ( PIP3_AKT-t308_s473 ( active AKT ) , 4E-BP_tot , MAPK* , S6K_tot and protein ) were normalized by dividing the obtained concentration by the value obtained from the original parameter model . These normalized fold change were plotted against logarithmic value of the parameter scale factor to obtain the sensitivity plots . | Memory formation involves the controlled production of new proteins close to the site of input stimuli on nerve cells . Strong inputs , in combination with growth factors , stimulate the synthesis of several kinds of synaptic proteins . These new proteins are believed to participate in remodeling the contacts between cells . This gives rise to a potentially unstable situation of a self-modifying cellular machine , because the new proteins rebuild their own inputs and their own production machinery . We have analyzed these interactions by modeling multiple inputs and the process of self-modifying feedback . We find that runaway modifications are prevented in two ways: first , a molecule called mTOR acts as a gate to suppress synthesis except under very tightly regulated conditions . Second , the feedback processes operate in a range where it is very unlikely that they can give rise to runaway buildup . Thus , the system avoids instability even though it is capable of modifying itself in response to many kinds of inputs . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
]
| [
"computational",
"biology/computational",
"neuroscience",
"neuroscience/neural",
"homeostasis",
"neuroscience/neuronal",
"signaling",
"mechanisms",
"computational",
"biology/signaling",
"networks",
"computational",
"biology/systems",
"biology"
]
| 2009 | Signaling Logic of Activity-Triggered Dendritic Protein Synthesis: An mTOR Gate But Not a Feedback Switch |
Collective decision making and especially leadership in groups are among the most studied topics in natural , social , and political sciences . Previous studies have shown that some individuals are more likely to be leaders because of their social power or the pertinent information they possess . One challenge for all group members , however , is to satisfy their needs . In many situations , we do not yet know how individuals within groups distribute leadership decisions between themselves in order to satisfy time-varying individual requirements . To gain insight into this problem , we build a dynamic model where group members have to satisfy different needs but are not aware of each other's needs . Data about needs of animals come from real data observed in macaques . Several studies showed that a collective movement may be initiated by a single individual . This individual may be the dominant one , the oldest one , but also the one having the highest physiological needs . In our model , the individual with the lowest reserve initiates movements and decides for all its conspecifics . This simple rule leads to a viable decision-making system where all individuals may lead the group at one moment and thus suit their requirements . However , a single individual becomes the leader in 38% to 95% of cases and the leadership is unequally ( according to an exponential law ) distributed according to the heterogeneity of needs in the group . The results showed that this non-linearity emerges when one group member reaches physiological requirements , mainly the nutrient ones – protein , energy and water depending on weight - superior to those of its conspecifics . This amplification may explain why some leaders could appear in animal groups without any despotism , complex signalling , or developed cognitive ability .
Social animals have to coordinate their activities in order to maintain the advantages of group living [1]–[3] . This coordination constitutes one of the major challenges of any animal society , including human beings , and arouses the interest of scientists , sociologists , and politicians [4]–[9] . Whatever the group size or the level of communication – global or local [8] , [9] – several categories of group decision making have been described: a leadership process where one individual will propose or impose a decision that other group members will follow [10]–[15] , and a voting process in which each individual indicates a direction , for instance , and after which the group will move in the direction of the majority [4] , [16]–[18] . A group leader is usually defined as the individual initiating group movements but also as the individual coordinating individual during the group progression , and then mainly at the front of the progression [4] , [8]–[13] . In different species of animals , leadership is not necessarily homogeneously distributed among group members [8]–[15] . Some individuals are more likely to become leaders thanks to specific internal or social traits increasing their probability of initiating a movement [10]–[12] . Studies of elephants [19] , ravens [20] , or fishes [21] have reported that some individuals may have a greater knowledge about their environment – which is the best site to eat or to drink – and these individuals have been observed leading the group more often than their conspecifics . In other species , individuals having a high social status , in terms of dominance or affiliation , also have a greater likelihood of being leaders . Probably the best known examples come from wolves and gorillas [10] , [22] where the dominant male or couple is described as always deciding for the entire group . In Tonkean macaques , however , the most affiliated individuals – who are not necessarily the most dominant ones – seem to have a greater influence than their conspecifics in collective decision making [23] . However , one of the major factors influencing leadership should be the different physiological requirements of group members [10]–[12] . Such heterogeneity implies conflicts of interests between individuals that must be resolved in order to maintain group cohesion . Leaving the leadership to highly motivated individuals seems to be one compromise . Indeed , the moving decision seems to be taken by those with highest needs in fishes , zebras , and primates [8] , [9] , [14] , [24] . Nevertheless , we still lack data on the way leaders emerge and the viability of the decision-making system concerning the entire group's satisfaction . Using a modelling approach , Rands et al . [12] , [25] and Conradt et al . [26] showed that individuals with the highest nutrient requirements can be more prone to lead the group if there is an advantage to foraging together . Their studies were however restricted to pairs of individuals or to situations in which individuals faced two mutually exclusive target destinations only . Here , we use a state-dependent dynamic model [27] , [28] to determine how nutrient and social requirements can determine the synchronization of a group of n individuals , their activity budget , and the emergence of leaders . This kind of models allows us to understand how simple rules based on nutrient requirements and social factors explain synchronization between group members [27] but also segregation as shown in ungulates [28] , where each individual has some requirements to satisfy ( nutrient requirements such as protein , energy and water but also other social requirements and resting ) . We assume that if there is an advantage to being in a group , then the group members should synchronize their activities in order to stay cohesive . We assume that individuals do not know the requirements of their conspecifics ( and further show that such ability may not be necessary for effective group coordination ) . Each individual requirement combines a reserve and a motivation that we call probability to lead . When the reserve decreases , the probability increases . At one moment , the individual with the lowest reserve – among its own needs and in comparison to other individuals – will have the highest probability to lead ( these individual probabilities are compared at each time step ) and will decide for the entire group on changing activity in order to fulfil its respective reserve [8] , [9] , [12] , [25] , [26] . In the next step , when the need of the previous leader is satisfied , a new leader will emerge and decide for the whole group . We applied this condition in our model in order to assess if this simple hypothesis “leading according to needs and deciding for all the group” is viable and if so , how the leadership will be distributed among group members less or more heterogeneous in their needs .
We first tested a group of two individuals , A and B , with two needs , x1 and x2 . We set three conditions: 1 ) needs are equal ( a1 = a2 = b1 = b2 ) , 2 ) needs of each individual are different but their sums are equal between individuals ( a1>a2; b1>b2; a1+a2 = b1+b2 ) and 3 ) the sum of needs for individual A are always superior to the one of individual B ( a1+a2>b1+b2 ) . Values of needs for each tested group of each condition are detailed in table 1 . We tested 10 groups for each condition . Results show that the decision system is viable , no individual dies , i . e . , no individual has needs not met ( reserves go to 0 ) , whatever the tested condition . When sums of needs are equal between individuals ( conditions 1 and 2 ) , the leadership ( proportion of decisions , i . e . , initiations per individual ) is equally distributed between the two individuals ( Fig . 1A ) , even if , at each time step , one individual is the leader and the other one is the follower according to the reserves' difference . This result is similar to the one of the paper of Rands et al . [25] where individuals are identical . However , when the sum of needs is superior for one individual ( Fig . 1B ) , this one becomes the leader of the pairs of individuals and the other individual becomes a follower almost all the time ( Kolmogorov-Smirnov test , P<0 . 0001 ) . The leadership difference between individuals increases with their relative difference of needs in a logarithmic way ( curve estimation test: R2 = 0 . 96 , P<0 . 00001; Fig . 1C ) . This result is similar to the one of Rands and colleagues [12]: leaders emerge when individual reserves differ . In a second step , we used data coming from animals in order to validate our model and to study emergence of leaders in larger groups . . According to needs of macaques , animals were divided into five categories: adult males , adult cycling females , lactating females , subadults , and juveniles . An individual has five requirements to satisfy: water , protein , energy , resting , and socializing [29]–[34] . The nutrient requirements of an individual ( water , protein and energy ) depend on its body mass whilst social and resting needs did not [31]–[34] . We chose to include social activity in the model because many social species spend time maintaining their relationships and group cohesion [31]–[35] . Group composition ( table 2 ) and individual characteristics ( table 3 ) are based on data on macaques and are detailed in the method section . We tested 10 groups of 5 , 10 and 20 individuals with same needs ( individuals of the same category and with the same body mass ) and 10 ones with different needs ( individuals of both different categories and different body masses ) . Simulations showed that the system – leadership by those in need – is sustainable in groups of 5 , 10 and 20 individuals . All individual requirements are satisfied at the end of simulations , whatever the group composition . Moreover , the group activity budget is fairly similar to the activity budget of wild primate groups ( 27 . 3±1 . 7% of time devoted to moving , 33 . 8±1 . 7% to foraging , 21 . 7±0 . 7% to resting , and 17 . 2±3 . 1% to socializing ) . All individuals could become leaders but the distribution of the leadership proportion is not the same according to the requirements' heterogeneity ( equal or different needs; Kruskal-Wallis test , P<0 . 001 ) . In groups with similar needs , the proportion of leadership differs weakly between individuals and is about 10% per individual . The relation between leadership proportion and rank ( i . e . , individuals were ranked from the most frequent leader to the less frequent one ) is linear ( linear curve estimation test: R2 = 0 . 92 , F1 , 8 = 93 . 05 , P<0 . 00001 , y = −0 . 0006x+0 . 1339 ) . The leadership is about 14% for the individual that decides the most and 7 . 6% for the individual that decide the least ( Fig . 2A ) . This result corresponds to the equiprobability of being leader per individual ( proportion divided by the number of individuals per group ) . On the other hand , the leadership is not equally distributed in heterogeneous groups . The relation between the proportion of leadership and individuals is exponential ( exponential curve estimation test: R2 = 0 . 83 , F1 , 8 = 38 . 07 , P = 0 . 0002 , y = 3 . 5727e−4 . 602x , Fig . 2B ) , with one individual being responsible for 38% to 95% of decisions per group , while some individuals decide only in 0 . 0003% to 0 . 0007% of cases per group . We obtained the same relationship with groups of 5 ( exponential curve estimation test; R2 = 0 . 97 , F1 , 3 = 12 . 81 , P<0 . 00001 , y = 0 . 9825x−3 . 207 , Fig . 3A ) and 20 individuals ( exponential curve estimation test; R2 = 0 . 96 , F1 , 18 = 498 . 95 , P<0 . 00001 , y = 11 . 48x−4 . 86 , Fig . 3B ) . We compared this unequally distributed leadership to the requirements of individuals in order to understand how so many differences can emerge in heterogeneous groups . We calculated the relative difference in requirements ( corresponding to the highest probability to lead ) between each leader and other individuals . The relationship between the leadership and this difference in requirements follows a sigmoid curve ( with a threshold S of 1 . 37 and a minimal n value of 30; curve estimation test: R2 = 0 . 71 , F1 , 108 = 269 . 72 , P<0 . 00001; Fig . 4A ) . The n value represents the sensitivity of the process . The higher the n value is , the more sensitive the process is ( quick and sudden transition between the two states ) . In our context , this means that one individual becomes the most frequent and prominent leader of a group as soon as one of its requirements exceed about 137% of those of one of its conspecifics . This transition between equally distributed leadership and one exclusive leader is highly non-linear , given the n value we observed . The same sigmoid law is observed between the proportion of leadership and the body mass of individuals ( sigmoid curve estimation test: R2 = 0 . 66 , F1 , 108 = 205 . 73 , P<0 . 00001; Fig . 4B ) . When the mass of an individual is more than 170% ( S = 1 . 7 , n = 30 ) of those of its conspecifics , this individual is the main group leader . Except for lactating females , requirements and then leadership are related to body mass in about 60% of cases . The rest of the decisions concern resting and socializing and are not related to mass . We obtain similar results for groups of 5 ( sigmoid curve estimation test: R2 = 0 . 71 , F1 , 53 = 111 . 52 , P<0 . 00001 , , Fig . 5A ) and 20 individuals ( sigmoid curve estimation test: R2 = 0 . 16 , F1 , 218 = 42 . 13 , P<0 . 00001 , , Fig . 5B ) . For 8 out of 10 groups of 20 individuals , 4 . 6±2 . 2 group members were never leader . They were satisfied by following their conspecifics .
Leading by those highest in need resembles the results obtained by Rands et al . [25] , where the individual with the lower reserve spontaneously becomes the leader . Moreover , a recent study by Conradt et al . [26] showed that a small minority of individuals with strong needs are more prone to lead the group than a larger majority of individuals with few needs . However , it is the first time that a threshold [2] , [18] , [36] has been demonstrated concerning the emergence of leadership . The decision-making system implies high differences in leadership proportion whilst relatively small differences are observed in the requirements of individuals . The threshold we obtained in this study is probably dependent on 1 ) the group structure of primates ( one or a small number of males compared to the other categories ) [35] and 2 ) to the physiology of primates [31]–[34] . Indeed , in primates , and especially in macaques , a sexual dimorphism exists and males may reach a mass 150 to 200% superior to the one of females . Several authors suggested that dominant individuals are the only leaders in several species [9]–[12] , [14] , [22] , [23] . The dominance is however strongly correlated to the body mass and then to the nutrient requirements of animals [10] . This indirect effect of dominance on leadership , through the needs and then the probability to initiate a movement , needs to be taken into account in subsequent studies testing dominance effects . For instance , two field studies in baboons showed that the main leader – the individual initiating most of movements – was the dominant male . However this male is also certainly the biggest individual in the group . In the study of Stueckle and Zinner [36] , the four males of the group , bigger than females , are the ones initiating the most of movements ( Fig . 6 ) . Moreover , the distribution of leadership also follows an exponential as the one in the study model . We may suggest that the slope of this exponential distribution of leadership will be less or more important according to the group composition . This slope would be around 0 when the group is homogeneous and increases with group heterogeneity . The non-linear differences in leadership among group members eventually emerge from two simple rules: individuals need to remain cohesive and the individual with the lowest reserve at one moment decides for the group [2] , [3] , [24]–[26] . Mechanisms of coordination and cohesion do not need complex signalling or complex cognitive ability [2] , [3] , [13] , [24] . The emergence of a unique leader may also occur when decisions are not necessarily imposed on other group members but because other individuals do not express the necessity to move or to make a decision . An individual becomes a leader because its conspecifics decide to follow it [8] , [9] . This outcome may make important contributions to our understanding of decision making in animal and human societies .
The model was developed in Netlogo 3 . 15 [37] . The model and model's procedures can be found in the supplementary material “Dataset S1” . One time-step in the simulation represents one minute . We defined the probability to lead α for the requirement A and the individual i as:The probability to lead for the individual i is:In this way , the probability to lead can vary between 1 ( highest probability to lead , weakest reserve ) and 0 ( weakest probability to lead , highest reserve ) . Each reserve is bounded by a maximum above which each group member cannot gain further reserves and a minimum at which each group member is assumed to die if it is reached . At each timestep ( equal to one minute ) , each reserve of each individual decreases ( i . e . , expenditure ) depending on the individual category and the current activity . This reserve decrease will increase the individual probability to lead . In order to fulfil this reserve , the individual should have to carry out the corresponding activity ( i . e . , intake ) . This gain may be done by becoming a leader or by following the leader . We implement optimal foraging decisions in the model: when an individual decides to forage , it will forage until its reserve has been fulfilled . After the end of each activity period , the individual with the highest probability to lead Pi becomes the new leader . Individuals have a walking speed of 0 . 4m . s−1 . The group environment is two-dimensional environment of 96×96 connected cells . Each cell represent one meter . Each cell has four immediate neighbours and the sides of the arena were joined to form a torus . The number of areas where animals fulfil their reserves is two for the first model with two individuals having two needs and four for the model from 5 to 20 individuals having five needs ( see details below ) . At the start of a simulation , individuals are at the same distance of each area ( i . e . , at the middle of the torus ) . According to the distribution of areas inside the torus , groups have a travel distance between two areas ranging from a minimum of 25 meters to a maximum of 75 meters . This range fits with travel distances in primate species of similar body mass and similar group size [4] , [38]–[41] . Positions of areas were fixed in our model but this does not affect results since variability among needs – what is the highest need and the weakest one – is much more important between individuals and groups . This means that the areas corresponding for instance to the two highest needs for an individual are not always the closest ones . There is no intragroup competition in this model: all individuals can occupy the same area . We run 1000 simulations for each group . A simulation stops when one reserve of one individual reaches 0 or after 90 days . The two individuals have two needs and thus two daily requirements . Values of these requirements for each condition and each individual are described in table 1 . We tested ten different groups for each condition . Expenditures of each reserve are 0 . 07±0 . 035 units . min−1 . Intakes are 10 units . min−1 . The environment is composed of two areas , one for each requirement . Individuals have to move to the respective area to fulfil each reserve . According to data in macaques , the daily protein requirement is estimated to 2 . 54g . day−1 . kg−1 , daily energy requirement to 351 . 7Kcal . day−1 . kg−1 , and daily water requirement to 0 . 24ml . KJ−1 , except for lactating females for which these requirements are higher than the ones of non lactating females ( about 125% for proteins and 200% for energy and water of requirements of non lactating females ) [31]–[34] . Social and resting times are not dependent on body mass . Individual expenditure per need and activity is described in table 4 . Details about individual intake rate per need are in table 5 . The environment is composed of four areas: one area for foraging for proteins , one area for foraging for energy , one waterhole , and one resting site [42] . When individuals need energy , proteins , or water , they have to move toward the respective areas . Until the group is in a specific activity among the five ones ( eating proteins , eating energy , drinking water , resting or socializing ) , each individual gains a certain amount of the requirement according to its category ( table 5 ) . Concerning resting , individuals need to go to the resting site for the night ( at the 720th time-step and for 720 time-steps ) , but during the day they can rest in any area . The same rule applies to socializing . Concerning resting and socializing activity , we fixed a minimal period of 5 minutes for doing these activities . Differences in leadership between individuals were tested using a Kolmogorov-Smirnov test for groups of 2 individuals and a Kruskall-Wallis test for groups from 5 to 20 individuals . The relations between the proportion of leadership and differences in needs or mass were determined through a curve estimation test . We compared observed curves to exponential , linear and sigmoid ones . Only theoretical curves best fitting with observed data are indicate in results . Analyses were performed in SPSS 10 . 00 . α was set at 0 . 05 . Means were ± S . E . M . | Making decisions together to reach a consensus is one of the most important challenges of any society . In some communities , however , some leaders have more weight in the decisions than the other individuals . Similar rules exist in animal societies . Studies on animal groups have shown that some individuals are more likely to be leaders because of their social power or the pertinent information they possess . This leader may be the dominant one , the oldest one , but also the one having the highest physiological need . However , how may other group members have their needs satisfied if always the same individual decides ? To gain insight into this problem , we build an agent-based model where group members have to satisfy different needs but the individual with the lowest reserve decides when and where to move for all its conspecifics . This simple rule leads to a viable decision-making system that satisfies all individuals and suits their requirements . However , a single individual , the one with the highest needs , becomes the leader in 38% to 95% of cases according to the heterogeneity of needs in the group . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"ecology/evolutionary",
"ecology",
"ecology/behavioral",
"ecology",
"evolutionary",
"biology/animal",
"behavior",
"computational",
"biology/evolutionary",
"modeling",
"computational",
"biology/systems",
"biology"
]
| 2010 | Differences in Nutrient Requirements Imply a Non-Linear Emergence of Leaders in Animal Groups |
The helminth parasite Fasciola hepatica secretes cathepsin L cysteine proteases to invade its host , migrate through tissues and digest haemoglobin , its main source of amino acids . Here we investigated the importance of pH in regulating the activity and functions of the major cathepsin L protease FheCL1 . The slightly acidic pH of the parasite gut facilitates the auto-catalytic activation of FheCL1 from its inactive proFheCL1 zymogen; this process was ∼40-fold faster at pH 4 . 5 than at pH 7 . 0 . Active mature FheCL1 is very stable at acidic and neutral conditions ( the enzyme retained ∼45% activity when incubated at 37°C and pH 4 . 5 for 10 days ) and displayed a broad pH range for activity peptide substrates and the protein ovalbumin , peaking between pH 5 . 5 and pH 7 . 0 . This pH profile likely reflects the need for FheCL1 to function both in the parasite gut and in the host tissues . FheCL1 , however , could not cleave its natural substrate Hb in the pH range pH 5 . 5 and pH 7 . 0; digestion occurred only at pH≤4 . 5 , which coincided with pH-induced dissociation of the Hb tetramer . Our studies indicate that the acidic pH of the parasite relaxes the Hb structure , making it susceptible to proteolysis by FheCL1 . This process is enhanced by glutathione ( GSH ) , the main reducing agent contained in red blood cells . Using mass spectrometry , we show that FheCL1 can degrade Hb to small peptides , predominantly of 4–14 residues , but cannot release free amino acids . Therefore , we suggest that Hb degradation is not completed in the gut lumen but that the resulting peptides are absorbed by the gut epithelial cells for further processing by intracellular di- and amino-peptidases to free amino acids that are distributed through the parasite tissue for protein anabolism .
Fasciolosis is a disease caused by helminths of the genus Fasciola . F . hepatica is found in temperate climates whereas F . gigantica is predominant in tropical regions . However , the distribution of the two species overlap in the Asia-Pacific region where hybrid forms have been isolated [1] , [2] . Fasciolosis is of major global economic importance as it infects primary production livestock of humans , especially sheep , cattle and water buffalo . Moreover , epidemiological surveys carried out over the last 15 years have uncovered fasciolosis as a significant human zoonosis . To-date , high prevalence of human infection has been reported in South America ( Ecuador , Peru and Bolivia ) , Vietnam , Thailand , Egypt and Iran [2] , [3] . Animals and humans become infected by ingesting vegetation contaminated with infective larvae which emerge from cysts and migrate through the intestinal wall and liver tissue causing extensive tissue damage and haemorrhaging as they burrow and feed . The parasites then enter the bile ducts where they mature and produce eggs that are carried into the environment with the bile juices [1] . The success of F . hepatica as a parasite is related its ability to infect and complete its lifecycle in wide range of mammalian hosts . Besides domestic ruminants and humans these include a large number of relevant reservoir hosts , such as deer , rabbits , hares , rats and mice [2] . Over the last few centuries European colonisation accelerated the distribution of the disease by introducing infected animals into many countries [2] , [3] . Most remarkably , within this relatively brief time period the parasite has adapted to local host species such as camelids in Africa , llamas and alpaca in South America and kangaroo in Australia [2] . Fasciola parasites infect and survive in their hosts by secreting cathepsin cysteine proteases . RNAi-mediated knock-down of cysteine protease activity of infective larvae was shown to prevent their ability to migrate through the host intestinal wall [4] . Also , blocking the function of these enzymes using anti-cysteine protease inhibitors or by vaccination with purified enzymes protects animals from infection [5] , [6] . The primary function of the F . hepatica cathepsin L proteases is in the digestion of host haemglobin ( Hb ) , the main source of nutrient for the parasite . This takes place within the lumen of the parasite gut , which is believed to be slightly acidic at around pH 5 . 5 [7] , [8] . Adult parasites draw blood with a muscular pharynx through punctures they make in the wall of the bile duct and use it to supply the amino acids needed for the massive production of eggs [7] . Due to the blind-ended nature of the adult parasite gut it must be emptied regularly by regurgitation ( approximately every three hours ) and refilled with fresh blood [9] . This process is also important for the extrusion of the cathepsin L proteases into the host tissues where they are involved in additional pivotal functions to parasitism including penetration of the host's tissues , cleavage of host immunoglobulins and suppression of immune cell proliferation [3] , [6] . These extracorporeal functions of the parasite enzymes are performed , therefore , in an environment of neutral physiological pH . The functions of proteases are not only related to their physico-biochemical properties but also to their cellular/tissue location and physiological environment ( particularly pH ) . Since F . hepatica cathepsin L proteases are required to function both inside and outside the parasite we considered it important to investigate the regulatory influence of pH on the autocatalytic processing and activation of the inactive zymogen , and on the structural stability and hydrolytic activity of the major secreted enzyme cathepsin L1 ( FheCL1 ) . We found that this enzyme was most rapidly activated at low pH but , once activated , was stable and functional over a broad pH range with an optimal hydrolytic activity at pH 6 . 2 . While FheCL1 readily cleaved peptide and protein ( ovalbumin ) substrates at neutral pH , Hb was resistant to cleavage at this pH . The degradation of Hb required acid-induced structural changes that made it susceptible to FheCL1 cleavage . Degradation was enhanced by the presence of small thiol agents , such as glutathione and cysteine , which activate FheCL1 and are present in physiologically-relevant concentrations in red blood cells and plasma . Our experiments suggest that Hb is digested by the parasite in a microenvironment , likely between the lamellae of the gut epithelial cells , at a pH of approximately 4 . 5 . Under low pH and reducing conditions FheCL1 is capable of generating small peptides but not free amino acids . We propose that these peptides are absorbed by the gut epithelial cells of the parasite where further processing takes place by intracellular dipeptidases [10] and aminopeptidases [11] to release amino acids that are distributed to the parasites tissues and used for protein anabolism .
Z-Phe-Arg-NHMec was obtained from Bachem ( St . Helens , UK ) . E-64 , DTT , l-cysteine , GSH ( reduced glutathione ) , EDTA and ovalbumin were obtained from Sigma-Aldrich ( Sydney , Australia ) . Prestained molecular mass markers were obtained from Invitrogen ( Victoria , Australia ) . Expression , production and purification of recombinant wild-type proFheCL1 and variant proFheCL1Gly25 ( procathepsin L ) in the yeast Pichia pastoris have been described elsewhere [12] , [13] . The variant proFheCL1Gly25 is an inactive zymogen since the active site Cys was replaced by a Gly . Auto-activation of the active wildtype proFheCL1 was carried out by incubating 0 . 2 mg/ml enzyme at 37°C in 100 mM sodium acetate buffer , pH 4 . 5 , containing 1 mM DTT and 1 mM EDTA . Aliquots ( 15 µl ) were removed at time intervals and added to tubes containing 1 µl of 1 mM E-64 to stop the reaction . Proteolytic cleavage of the prosegment was visualised by 15% SDS-PAGE . Auto-activation was also monitored in the presence of the fluorogenic substrate Z-Phe-ArgNHMec by measuring the release of fluorescence over time using a KC4 Bio-Tek micro-plate reader in 96-well fluorescent plates . proFheCL1 ( 5 nM ) was incubated in 100 mM buffer pH 4 . 0–pH 7 . 0 in the presence of 2 µM Z-Phe-Arg-NHMec . Final linear rates of substrate hydrolysis at each pH were measured with FheCL1 after auto-catalysis was completed . Fluorescence assays measuring activity of mature FheCL1 was carried out in 96-well plates using a KC4 Synergy HT micro-plate reader ( Bio-Tek Instutments Inc . , Vermont , USA ) . Assays were carried out with a final substrate concentration of 0 . 5 µM in a volume of 200 µl . When [S]<KM the initial rate is proportional to kcat/Km . Assays contained 0 . 14 nM cathepsin L1 in the following buffers: 100 mM formate ( pH 3 . 24–4 . 0 ) , 100 mM sodium acetate ( pH 4 . 0–5 . 5 ) , 100 mM sodium phosphate ( pH 5 . 5–8 . 0 ) , and 100 mM sodium borate ( pH 8 . 0–10 . 0 ) . The assay also contained final concentrations of 1 mM DTT and 1 . 0 mM EDTA . The data were fitted to the equation: Stability of FheCL1 was investigated by incubating 0 . 1 mg/ml enzyme in 100 mM buffer ( pH 2 . 5–pH 9 . 0 ) at 37°C . Enzyme activity towards 5 µM Z-Phe-Arg-NHMec in 100 mM sodium acetate buffer , pH 5 . 5 and containing 1 mM DTT was measured at time intervals over a 10-day period . Human red blood cells were washed three times by resuspending 0 . 25 ml of whole blood in 5 ml PBS and centrifugation at 5000 rpm . The supernatant with the buffy coat was removed each time . After the final wash , the cells were lysed to release haemoglobin ( Hb ) by adding 1 ml ice-cold distilled H2O for 10 min and then the suspension was centrifuged at 15000 rpm to remove insoluble material [14] . To remove any free amino acids or low molecular mass material Hb was dialysed twice against 1 . 5 L phosphate-buffered saline ( PBS ) , pH 7 . 3 , for 3 h using a dialysis membrane with a 3000 Da molecular mass cut-off ( Sigma Chemical Co . , Sydney , Australia ) . Hb was quantified using an extinction coefficient of 125 000 M−1 cm−1 at 414 nm [15] and was in good agreement with the total protein in lysates measured by the Lowry method [16] using BSA as standard . Spectrophotometry was carried out in in 96-well plates in a KC4 Synergy HT micro-plate reader . Hb was diluted to a final concentration of 5 µM into 100 mM buffer and denaturation was recorded for one hour by monitoring the decrease in absorbance at 414 nm [15] . Buffers used were 100 mM sodium acetate buffer , pH 3 . 5–pH 5 . 5 and 100 mM sodium phosphate buffer , pH 6 . 0–pH 7 . 0 . Absorption spectra were recorded after one hour for each sample from 600 nm–300 nm . Hb denaturation was also monitored for 90 minutes at pH 4 . 5 and pH 7 . 0 in the presence of 1 mM GSH , with or without FheCL1 . Stock FheproCL1Gly25 in PBS was dialysed into 50 mM sodium phosphate buffer , pH 7 . 5 or 50 mM sodium acetate buffer , pH 4 . 0 , to remove any NaCl that could interfere with the CD spectrum . CD spectra of 5 . 3 µM proFheCL1Gly25 ( ∼0 . 2 mg/ml ) were recorded over the wavelength range 195–250 nm , in steps of 0 . 5 nm and speed of 20 nm/min using a Jasco720 spectropolarimeter in quartz cuvettes with a 0 . 1 cm pathlength . Spectra were the average of three scans and were buffer baseline corrected . Hb ( 1 . 8 nmoles ) and ovalbumin ( 1 . 2 nmoles ) were incubated with FheCL1 ( 0 . 18 nmoles ) in 0 . 1 M buffers , pH 3 . 5–8 . 0 and containing 1 mM DTT . Control experiments contained no enzyme . The buffers used were 100 mM sodium acetate ( pH 3 . 5–5 . 5 ) and 100 mM sodium phosphate ( pH 5 . 5–8 . 0 ) . The reactions were stopped after 30 min by adding 1 µl 1 mM E-64 to the tube and aliquots were analysed by 15% SDS-PAGE under reducing conditions . Gels were stained with a 0 . 1% w/v solution of Coomassie Brilliant Blue R-250 in 40% methanol/10% acetic acid [17] . Hb ( 1 . 8 nmoles ) was digested with purified recombinant FheCL1 ( 0 . 9 nmoles ) in 0 . 1 M sodium acetate buffer ( pH 4 . 0 ) containing 1 mM GSH and 1 mM EDTA for 0 , 10 , 20 , 30 , 45 , 60 , 75 , 90 , 120 and 180 minutes at 37°C . 10 µl aliquots of the digests were analysed using NuPage Novex 4–12% Bis-Tris gels ( Invitrogen ) according to the manufacturer's instructions . Gels were stained with Colloidal Coomassie Blue G250 ( Sigma ) . Hb digests were spun at 13 , 000 rpm for 15 min to remove particulates and were concentrated to a final volume of 15 µl using a Concentrator 5301 ( Eppendorf ) . Peptides were analysed by nanoLC-ESI-MS/MS using a Tempo nanoLC system ( Applied Biosystems ) with a C18 column ( Vydac ) coupled to a QSTAR Elite QqTOF mass spectrometer running in IDA mode ( Applied Biosystems ) . Peak list files generated by the Protein Pilot v1 . 0 software ( Applied Biosystems ) were exported to local MASCOT ( Matrix Science ) and PEAKs ( Bioinformatics Solutions Inc . ) search engines for protein database searching . MS/MS data was used to search 3239079 entries in the MSDB ( 20060809 ) database using MASCOT whereas PEAKs software was used to search a custom-made database containing only human Hb-alpha and Hb-beta sequences . The peptide mass tolerance was set at 0 . 1 Da , oxidation of methionine residues was set as a variable protein modification and the “no enzyme” function was selected . For MASCOT searches , matches with a MOWSE score >70 were considered to be significant [18] , [19] and matched peptides achieving a score >60% were accepted during PEAKs searches . The matching peptides were then mapped onto the primary amino acid sequences of human Hb-alpha and Hb-beta to identify FheCL1 cleavages sites . For matrix-assisted laser desorption ionisation time-of-flight mass spectrometry ( MALDI-TOF MS ) the Hb digest was desalted and concentrated by zip-tip ( Millipore Perfect Pure C18 ) and spotted using 1 µL matrix ( α-cyano-4-hydroxycinnamic acid , 4 mg/mL in 70% v/v acetonitrile , 0 . 06% v/v TFA , 1 mM ammonium citrate ) onto a target plate , and allowed to air dry ( Australian Proteome Analysis Facility , Macquarie University , Sydney . ) . The sample was then analysed using a 4700 Proteomics System TOF mass spectrometer ( Applied Biosystems , USA ) operated in reflectron mode in the mass range of 100 m/z to 400 m/z . Spectra were analysed manually and externally calibrated using ACTH ( fragment 18–37 ) , neurotensin , angiotensin I , bradykinin to give a mass accuracy 50 ppm or less .
The zymogen of the F . hepatica cathepsin L1 , proFheCL1 ( Mr ∼38 kDa ) is auto-catalytically processed at pH 4 . 5 by inter-molecular cleavage and removal of the prosegment to release a fully mature and active enzyme ( Mr ∼25 kDa ) ( Figure 1A ) . Analysis of the in vitro auto-activation process by 4–20% SDS-PAGE shows that a band corresponding to the processed ∼25 kDa mature enzyme is observed within 5 minutes and that full removal of the prosegment from the zymogen occurs between two and three hours . Peptides representing products of the cleaved prosegment are observed below the 10 kDa molecular size marker ( Figure 1A ) . The rate of formation of an active mature enzyme from the inactive zymogen ( 5 nM ) was monitored between pH 4 . 0–7 . 0 by performing the autocatalytic reaction in the presence of the fluorogenic substrate Z-Phe-Arg-NHMec and calculating the rate of hydrolysis ( Figure 1B ) . The rate of hydrolysis of Z-Phe-Arg-NHMec , and hence the rate of activation from proFheCL1 to FheCL1 , was linear over this pH range; however , hydrolysis at pH 4 . 0 was ∼40-fold greater than at pH 7 . 0 indicating that autocatalytic activation occurs much more rapidly in an acidic environment ( Figure 1B ) . The relationship between the activity of the fully processed mature FheCL1 and pH was examined by determining the kcat/Km against Z-Phe-Arg-NHMec at various pH values in the range 2–10 . The results show that the enzyme has the capacity to cleave substrates over a wide pH range ( pH 3 . 0–9 . 0 ) . Maximal activity was observed between pH 5 . 5–7 . 0 with a peak at pH 6 . 2 ( Figure 1C ) . We recently described the production of a catalytically inactive proFheCL1 zymogen by replacing the active site Cys25 with a Gly [12] , [13] . This variant , proFheCL1Gly25 , was correctly folded but inactive and , therefore , unable to autocatalytically activate even at low pH . In the present study , we used this proFheCL1Gly25 variant to examine the stability of the zymogen at various pH values by subjecting it to analysis by circular dichroism ( CD ) in various solutions buffered in the pH range 4 . 0–7 . 5 ( Figure 2A ) . No significant difference in the far-UV CD spectra of proFheCL1 was observed showing that no conformational shifts occur in the secondary structure over this pH range . These data indicate that proFheCL1 would remain stable during the auto-catalytic activation process to FheCL1 , even at pH 4 . 0 . To investigate the susceptibility of the mature activated enzyme to pH denaturation , mature FheCL1 was incubated for at various time-periods at 37°C in buffers over the pH range 2 . 5–9 . 0 and then assayed for activity towards Z-Phe-Arg-NHMec in the presence of 1 mM DTT ( Figure 2B ) . The enzyme exhibited optimal stability at pH 4 . 5; even following a 10-day incubation period the enzyme retained ∼45% activity at pH 4 . 5 and ∼5% activity at pH 3 . 0 demonstrating that FheCL1 is very stable in a moderately acidic environment . When incubated at pH 2 . 5 , enzyme activity was not completely lost until day three . The activity of cysteine proteases is enhanced in the presence of small thiol molecules that reduce the active site cysteine . Dithiothreitol ( DTT ) is typically included in reactions carried out in the laboratory , but since this is not a physiological-relevant thiol we investigated whether reduced glutathione ( GSH ) and cysteine could activate the mature FheCL1 at concentrations found in blood ( GSH is found predominantly in red blood cells at concentrations of approximately 1 . 2 mM , while cysteine is found in plasma at 0 . 23 mM , [20] ) . To do this FheCL1 was incubated for 5 minutes at pH 4 . 5 in a range of concentrations of dithiothreitol ( DTT ) , GSH and L-cysteine . Substrate ( Z-Phe-Arg-NHMec ) was then added and endopeptidase activity determined by monitoring release of -NHMec with time . We found that in the presence of DTT , GSH and L-cysteine FheCL1 exhibited similar activation curves with maximal enzyme activity observed in the presence of each reducing agent at a concentration of ∼0 . 1 to 1 . 0 mM ( Figure 3 ) . Since Hb is a major physiological substrate for FheCL1 we examined the pH dependence for its hydrolysis by FheCL1 and compared this to the hydrolysis of ovalbumin . Firstly Hb was incubated alone in solutions buffered at various pHs in the range 3 . 5 to 8 . 0 for one hour and then analysed by SDS-PAGE . We observed that in the pH range 5 . 0–8 . 0 the molecule migrated as a major band at ∼15 kDa representing the Hb-alpha and Hb-beta monomers and a minor band at ∼30 kDa representing the alpha-beta dimers . However , at lower pH values the intensity of the band at ∼30 kDa increased and new bands ≥50 kDa were observed most likely due to aggregation of Hb ( with incubation times of greater than one hour precipitation of the Hb was observed in the acidic pH solutions ) ( Figure 4A ) . When Hb ( 50 µg ) was incubated in the presence of FheCL1 ( 1 µg ) no difference was observed in the migration pattern within the pH range 5 . 0 to 8 . 0 ( compare Figure 4B with 4A ) . At pH 4 . 5 , however , addition of FheCL1 caused the bands at ∼15 kDa and ∼30 kDa to disappear and smearing in the respective lanes indicated the presence of low molecular mass products due to Hb digestion ( Figure 4A and B ) . These Hb bands underwent greater degradation by FheCL1 in reactions carried out pH 4 . 0 and 3 . 5 ( Figure 4A and B ) . The above results indicate that FheCL1 cannot digest Hb at pH≥5 , whereas digestion is efficient in acidic conditions of pH≤4 . 5 . The lack of digestion at pH≥5 is not due to the inability of FheCL1 to function in this pH range as the studies above showed that the enzyme could cleave peptide substrates optimally between pH 5 . 5 and 7 . 0 . To support this observation we analysed the digestion of the protein ovalbumin by FheCL1 over the pH range 3 . 5 to 8 . 0 . Ovalbumin incubated in various buffered solutions at pH values from 3 . 5 to 8 . 0 migrates in SDS-PAGE as a single band at ∼45 kDa ( Figure 4C ) . When ovalbumin was incubated with FheCL1 a series of digestive products ( <45 kDa ) were produced ( Figure 4D ) . SDS-PAGE clearly shows that degradation was optimal between pH 5 . 0 and 7 . 0 ( Figure 4D ) , which is in agreement with the optimal activity of the enzyme determined against the peptide substrate ( Figure 1C ) . To investigate the effect of pH on the structure of Hb we obtained absorption spectra of the molecule in various buffered solutions . The absorption spectrum of Hb at physiological pH is characterised by a large Soret peak 414 nm due to the bound heme moiety; disruption of the Hb conformation causes shifts in this peak ( Figure 5A ) . No alteration in the Soret peak was observed between pH 7 . 0 and pH 5 . 5 , but the height of the peak began to decrease at pH 4 . 5 . When Hb was exposed to pH 4 . 0 and pH 3 . 5 the Soret peak completely disappeared ( Figure 5A ) indicating that structural changes are occurring in the Hb molecule such that it can no longer bind the heme moiety [15] . The denaturation of Hb at low pH was shown to be a time-dependent process ( Figure 5B ) . Progress curves obtained by monitoring the decrease in absorbance at 414 nm clearly show that Hb is stable at pH 7 . 0 and 5 . 5 but that partial denaturation occurs at pH 4 . 5 . Hb denaturation was complete at pH 4 . 0 and 3 . 5 within one hour . To determine if the rate of Hb denaturation made it more susceptible to FheCL1 degradation we mixed Hb at pH 7 . 0 and pH 4 . 5 in the presence and absence of 5 µM FheCL1 and 1 mM GSH and monitored denaturation at 414 nm for 1 hour ( Figure 4C ) . The results show that while FheCL1 had no effect on Hb denaturation at pH 7 . 0 , the rate of Hb denaturation/digestion was significantly increased at pH 4 . 5 in the presence of the protease . Thus the Hb molecule at physiological pH is resistant to proteolysis by FheCL1 but at pH 4 . 5 alterations in its structure take place that make it susceptible to hydrolysis , which is consistent with our SDS-PAGE analysis described above ( Figure 4 ) . Finally , the reducing agent GSH alone , at a concentration of 1 mM , had no effect on Hb denaturation at pH 4 . 5 or pH 7 . 0 ( Figure 5C ) . To examine the process of Hb degradation by FheCL1 Hb was mixed with the protease at pH 4 . 0 for 120 minutes at 37°C . Reactions were stopped at several time points by addition of E-64 ( an irreversible inhibitor of cysteine proteases ) and the degradation products were analysed by SDS-PAGE ( Figure 6A ) . The bands representing the 15 kDa Hb monomers and 30 kDa Hb dimers were gradually degraded to smaller protein bands in the molecular size region of 3–10 kDa within the first 10–20 minutes of the reaction and completely degraded between 60 and 120 min . It is noteworthy that during this digestive process the FheCL1 ( ∼25 kDa ) was not degraded ( Figure 6A ) supporting our earlier data showing that the enzyme is very stable under acid conditions ( Figure 2 ) . To identify the cleavage sites for FheCL1 within Hb , the 10 min and 120 min reaction aliquots were analysed by mass spectrometry . The peptides were then mapped onto the primary amino acid sequences of human Hb-alpha and Hb-beta to identify FheCL1 cleavage sites . Within 10 mins FheCL1 cleaved Hb-alpha at 47 sites and Hb-beta at 52 sites while at 120 min additional cleavage sites , totalling 83 sites in Hb-alpha and 89 sites in Hb-beta were observed ( Figure 6B ) . Examination of the cleavage map presented in Figure 6B shows that within a 10 min time-frame FheCL1 could generate small peptides of 4–8 amino acids from Hb . The map also indicates that these would conceivably be further degraded to release dipeptides and free amino acids after 120 mins . The 120 min digests were , therefore , analysed by LC-MS/MS to determine the masses and sequence identities of the resulting hydrolytic products . This analysis revealed that FheCL1 had degraded Hb into peptides ranging from 3–26 amino acids in length ( Figure 7 ) but not dipeptides or free amino acids . The average length of the released peptides ( from both the Hb alpha and beta chains ) was 10 amino acids with 13- and 12-residue peptides occurring most frequently in the digested Hb alpha and beta chains , respectively . Accordingly , FheCL1 must not cleave all Hb molecules in the same manner and , thus , the cleavage map shown in Figure 6B represents a composite of cleavage sites . To verify that free amino acids and/or small peptides ( di- or tri-peptides ) were not end-products of the proteolysis the digests were also analysed by MALDI-TOF MS ( specifically using the mass range 100 m/z to 400 m/z ) . Only 12 mass ions were detected within this range the masses of which five could be mapped to di- or tri-peptides present in either the Hb-alpha or -beta chains . Importantly , ion masses corresponding to free amino acids were not observed . Residues present at the P2 position from the scissile bond interact with the S2 subsite of the active site of papain-like cysteine proteases and determine the efficiency by which the bond is cleaved [21] . Therefore , we examined the frequency of each amino acid in the P2 site of the proteolytic cleavage site identified in aliquots of the 10 min Hb digest described above ( Figure 8 ) . Consistent with our previously published studies using fluorogenic peptide substrates and positional-scanning of synthetic combinatorial libraries [13] FheCL1 preferentially cleaved bonds where the P2 position was occupied with hydrophobic residues; this preference followed the order Leu>Val>Ala>Phe , and was observed for the digestion of both Hb-alpha and Hb-beta ( Figure 8 ) . Due to the promiscuity of the FheCL1 for peptide bonds no obvious trend for P2 preference could be discerned in digests taken at 75–120 minutes ( data not shown ) . Finally , in support of other studies using synthetic combinatorial libraries [13] the P1 position could be occupied by many amino acids but most preferentially Leu .
Proteomic analysis of proteins secreted from adult F . hepatica parasites in situ within the bile ducts [22] and in culture [22] , [23] showed that >80% of the secreted proteins are cathepsin L cysteine proteases . Furthermore , no other class of endopeptidase or exopeptidase was identified in these secretions demonstrating the exclusive reliance of the mature parasites on cathepsin Ls [22] , [23] . The cathepsin L proteases are synthesised within epithelial cells lining the parasite gut; these cells have both a secretory and absorptive function and spread extended lamellae into the gut lumen [24] . We have shown that cathepsin L zymogens are concentrated and stored in numerous secretory vesicles that lie at the apex or luminal side of these cells ready for secretion into the gut [7] , [24] . By the time the enzymes are secreted outside the parasite they have undergone complete processing to mature enzymes by removal of the prosegment portion , which informed us that the activation process takes place within the gut lumen [12] . The gut lumen of F . hepatica , like that of other trematodes such as the schistosomes , is believed to be maintained at a slightly acidic pH , approximately 5 . 5 [8] , [24] , [25] . In the present study we have demonstrated using recombinant pro-cathepsin L that auto-catalytic processing and activation can take place at neutral pH but that this occurs far more rapidly at lower pH values ( activation at pH 4 . 0 was 40-times faster than at pH 7 . 0 ) . Circular dichroism studies showed that the zymogen does not undergo any significant conformational alteration in the pH range 4 . 0 to 7 . 0 , and like the lysosomal cathepsin Ls of mammals is stable under acid conditions [26] . Enzymatic studies demonstrated that the mature activated enzyme is also very stable under the pH conditions it would experience in the parasite gut . We can conclude , therefore , that the slightly acidic conditions of the parasite gut are very suitable for the autocatalytic activation and digestive function of the cathepsin L proteases . The primary function of the cathepsin Ls in the parasite gut is to digest host macromolecules and tissues to usable products . Haemoglobin ( Hb ) is the principle source of amino acids for protein anabolism by the parasite and our present studies demonstrate that FheCL1 can efficiently degrade this substrate in an acidic environment , pH≤4 . 5 . Surprisingly , however , the cathepsin L protease could not cleave Hb at pHs≥5 . 0 , despite the fact that it has optimal activity towards small-peptide and protein ( ovalbumin ) substrates between pH 5 . 5 and pH 7 . 0 . These observations revealed the importance of low pH in regulating the structure of Hb and its susceptibility to proteolysis . By monitoring the Soret peak of Hb over a range of pH values we examined the conformational changes that are induced in the molecule . The Hb molecule retained it structure and bound heme in the pH range 5 . 5 to 7 . 0 but partial loss of heme-binding was observed when the pH was reduced to 4 . 5 . Hb underwent full denaturation after 1 hour at pH 3 . 0–4 . 0 , which in solution could be observed by precipitation of Hb in the reaction tubes . Thus , the susceptibility of Hb to FheCL1 proteolysis , as revealed by our SDS-PAGE analysis of digestion reactions , correlated with the pH whereby Hb becomes denatured . A recent study of acid-induced unfolding of Hb monitored by ESI-mass spectrometry proposed the following model for Hb denaturation:where subscripts “h” and “a” refer to holo- and apo-forms ( i . e . heme and non-heme forms of Hb , respectively ) [27] . This model indicates that the release of heme from the Hb molecule accompanies the separation and unfolding of the α and β subunits . The final steps in the denaturation scheme , from αβ dimers to heme-bound monomers and then to unfolded non-heme-binding monomers occurred at ∼pH 4 . 4 and ∼pH 4 . 0 , respectively . In our present study we showed that the addition of FheCL1 to Hb increased the rate by which Hb lost bound heme at pH 4 . 5 and confirmed that partial denaturation of Hb at this pH was sufficient to relax the structure of the molecule and make it susceptible to proteolysis . Our results are consistent with a much earlier study by Kimura et al . [28] who showed that pH-induced denaturation of Hb increased its susceptibility to trypsin digestion . In conclusion , our studies underscore the importance of the low pH of the parasite gut lumen for denaturing ingested Hb to facilitate its proteolytic hydrolysis . This process is not unlike the denaturation of proteins for hydrolysis in the acid human stomach . Determining the precise pH of the gut lumen presents a practical hurdle . As mentioned above the pH of the gut lumen in F . hepatica has been suggested to be ∼pH 5 . 5 [8] , [25] while that of the related trematode Schistosoma mansoni has been estimated to be pH 5 . 0–6 . 0 by Senft [29] , pH 6 . 0–6 . 4 by Chappell and Dresden [30] and 6 . 84 by Sajid et al . [31] . These were not direct measurements of the intraluminal pH but were generally obtained by measuring media into which parasites had extruded their gut contents . Our data showing that FheCL1 could not digest Hb at pHs≥5 . 0 is biochemical evidence suggesting that the site of proteolytic activity within the gut must be lower than pH 5 . 0 . Electron micrographs of the gut lumen of both F . hepatica [8] , [25] and S . mansoni [32] often visualise Hb as a dense precipitate , representing presumably denatured protein , in the vicinity of the gut lamellae . Halton's [8] , [25] interpretation of micrographs of the gut structure was that digestion in F . hepatica takes place between the lamellae of the secretory epithelial cells . Derived from these studies it was suggested that the pH in this local microenvironment is maintained at a more acidic pH than the gut lumen per se [24] . In support of this suggestion Delcroix et al . [33] found sequestered compartments between lamellae of the schistosome gut with pH as low as 3 . 9 . Their observation explains why the schistosome gut aspartic protease , SmCD , whose activity is confined to the range of pH 2 . 5–4 . 6 [34] could participate in Hb digestion . Similarly it explains the role of the schistosome cathepsin L cysteine protease , SmCL1 , which could efficiently cleave Hb only in the pH range 4 . 0–4 . 5 [14] . FheCL1 and other papain-like cysteine proteases are activated in the presence of low molecular mass thiols such as cysteine or DTT [35] , [36] . Although these compounds are routinely used to activate cysteine proteases as part of in vitro activity assays , they are not considered physiologically relevant reducing agents . GSH is the most abundant intracellular reducing agent and its concentration inside red blood cells is particularly high , estimated to be 1 . 192 mM by Mills and Lang [20] and ∼3 . 2 mM by Chappell et al . [37] . Here we found that GSH effectively enhances FheCL1 activity towards small synthetic substrates with an optimum at ∼0 . 1 to 1 . 0 mM GSH , and accelerates the digestion of Hb by FheCL1 . A concentration of ∼0 . 1 to 1 . 0 mM GSH could conceivably be reached in the parasite gut following lysis of ingested red blood cells notwithstanding variations in the size of the blood meal and dilution in the parasite gut . We used mass spectrometry to identify the cleavage sites of FheCL1 within Hb and to determine the size of the peptide products generated by its complete digestion . FheCL1 digested Hb at 83 cleavage sites in Hb-alpha and 89 sites in Hb-beta that resulted in short peptides of at least 4–14 amino acids , with some appearance of tripeptides . Residues in the P2 position are known to influence the efficiency of all papain-like cysteine proteases , and we found that those residues in Hb that were most susceptible to cleavage by FheCL1 were invariably a hydrophobic residue , and in the order Leu>Val>Ala>Phe . These results are consistent with our earlier studies using fluorogenic peptides and peptide libraries that showed FheCL1 to have a more restricted S2 active site compared to human cathepsin L and most readily accommodates hydrophobic P2 residues , particularly Leu [13] . It is pertinent to note that the amino acids Leu , Val , Ala , Phe make up approximately 42% of the Hb molecule and , therefore , we would propose that FheCL1 has been specificity adapted to degrade this substrate . However , our studies also show that cleavage by FheCL1 does not generate free amino acids and , by extension , suggests that Hb degradation is not completed within the parasite gut but that small peptides are taken up by the gut epithelial cells during their absorptive phase for further processing within cells [24] . The enzymes involved in this process likely include a dipeptidylpeptidase [10] and an aminopeptidase [11] that function at neutral pH and have been located by immunofluorescence microscopy within the cytosol of the epithelial cells . Our studies on F . hepatica point to a digestive machinery that requires proteases of only one mechanistic class i . e . cathepsin cysteine proteases . However , F . hepatica secretes different forms of these proteases with overlapping specificity that may complement each other [23] . Nevertheless , the mechanism of gut digestion appears to differ markedly from other helminths so far studied . Dalton et al . [38] were first to propose that schistosomes exploit a cascade involving aspartic and cysteine ( cathepsin B , L1 and L2 ) proteases within their gut lumen to achieve the complete degradation of Hb . The more recent studies by Delcroix et al . [33] , which exploited selective protease inhibitors and RNA interference ( RNAi ) to explore the mechanism of Hb digestion in schistosomes , supports the role of a network or combination of cysteine proteases , aspartic protease and an asparaginyl endopeptidase . However , Correnti et al . [39] showed that while knockdown of cathepsin B expression in schistosomes by RNAi retarded parasite growth it did not prevent Hb digestion in the parasite gut . On the other hand , Morales et al . [40] , also using RNAi , demonstrated that the cathepsin D aspartic protease is essential to survival of schistosomes through its pivotal role in Hb digestion . A multi-enzyme cascade involving cysteine and aspartic proteases is also necessary for Hb digestion in canine hookworm Ancylostoma caninum [41] and aspartic and cysteine proteases in the nematode Ostertagia ostertagi have also been shown to have activity against Hb [42] . Although several proteases appear to be involved in Hb digestion in these helminths it is still not clear whether digestion is regulated in an ordered manner , each enzyme working sequentially , or whether all proteases work simultaneously and in a random manner . Dalton et al . [7] , [24] suggested that the activity of each protease within the gut was regulated by pH , and therefore as the bloodmeal ( pH 7 . 0 ) was drawn into the gut the pH slowly decreased ( perhaps by proton pumps in the epithelial cells ) , each enzyme would come into play when its appropriate pH range for activity was reached; thus in schistosomes cathepsin B ( optimum pH 4 . 0–6 . 0 ) would be activated before cathepsin L ( optimum pH 4 . 0–4 . 5 ) , which would be followed by aspartic proteases ( optimum pH 2 . 9–4 . 0 ) . The cathepsin L proteases of F . hepatica also participate in functions outside the parasite gut; these include liver tissue degradation , cleavage of host antibodies and suppression of host immune cell function ( see [6] ) . The blind-ended gut of the parasite is emptied every 3 hours , thus depositing the cathepsin L proteases outside [9] . The extracorporeal roles of the proteases are performed at pH values that are between two and three pH units higher than the microenvironment at which the proteases function in the parasite gut . Our studies showing that FheCL1 are active and highly stable at neutral pH points to a specific adaptation of these molecules to carry out functions over a wide pH range . It is interesting to note that the pH optimum of the FheCL1 , pH 6 . 2 , is approximately mid-point between the pH values at which it works inside and outside the parasite . In contrast , lysosomal cathepsin Ls of mammals are active only at pHs values of approximately 4 . 5 , in keeping with the environment in which they function , and are inherently unstable at neutral pH so that cellular damage due to leakage from the lysosome is avoided [43] . To conclude , the helminth parasite F . hepatica secretes cathepsin L proteases that are specifically adapted to be functional at pHs at which they perform essential roles in this parasite's biology . The low pH of the parasite gut is important in regulating the activity of these proteases by providing a milieu whereby the proteases readily autocatalytically activate from inactive zymogens secreted by the surrounding epithelial cells , and by facilitating the denaturation of the protein substrates on which the proteases act . The mature cathepsin L proteases are extremely stable at this pH and their hydrolytic activity is greatly enhanced by GSH , most likely derived from ingested host red blood cells . FheCL1 is specifically designed to cleave peptide bonds with N-terminal hydrophobic residues which are most common in Hb with the goal to provide small peptides that can be absorbed by the gut epithelial cells for further processing to amino acids within cells before distribution to parasite tissues via amino acid transporters [24] . However , following completion of the digestive process in the gut lumen unwanted material is extruded which delivers the proteases to the outside where they can perform their additional extracorporeal roles at physiological pH conditions in which they are also highly active and stable . | Fasciola hepatica is a helminth parasite that causes liver fluke disease ( fasciolosis ) in domestic animals ( sheep and cattle ) and humans worldwide . Cathepsin L cysteine proteases ( FheCL ) are secreted by the parasite to invade its host , migrate through tissues and to degrade host haemoglobin ( Hb ) , a major source of nutrient to the parasite . FheCL1 is a very stable protease and active over a broad pH range ( 3 . 0–9 . 0 ) , making it very suitable for functions both inside and outside the parasite . The slightly acidic pH of the parasite gut not only regulates the autocatalytic activation of the proFheCL1 zymogen to an active FheCL1 protease but also induces relaxation of the Hb structure , making it more susceptible to proteolysis . The action of FheCL1 , which is enhanced by glutathione ( GSH ) , the major reducing agent found in red blood cells , degrades Hb to small peptides ( predominantly 4–14 residues ) that can be absorbed by the gut epithelial cells . Further processing within these cells by exopeptidases provides the necessary amino acids required for protein anabolism by the parasite . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
]
| [
"biochemistry/biocatalysis",
"biochemistry/protein",
"chemistry",
"infectious",
"diseases/helminth",
"infections",
"biophysics/protein",
"chemistry",
"and",
"proteomics",
"microbiology/parasitology"
]
| 2009 | The Importance of pH in Regulating the Function of the Fasciola hepatica Cathepsin L1 Cysteine Protease |
The current strategy for the elimination of onchocerciasis is based on annual or bi-annual mass drug administration with ivermectin . However , due to several limiting factors there is a growing concern that elimination of onchocerciasis cannot be achieved solely through the current strategy . Additional tools are critically needed including a prophylactic vaccine . Presently Ov-103 and Ov-RAL-2 are the most promising vaccine candidates against an Onchocerca volvulus infection . Protection induced by immunization of mice with the alum-adjuvanted Ov-103 or Ov-RAL-2 vaccines appeared to be antibody dependent since AID-/- mice that could not mount antigen-specific IgG antibody responses were not protected from an Onchocerca volvulus challenge . To determine a possible association between antigen-specific antibody responses and anti-larvae protective immunity in humans , we analyzed the presence of anti-Ov-103 and anti-Ov-RAL-2 cytophilic antibody responses ( IgG1 and IgG3 ) in individuals classified as putatively immune , and in infected individuals who developed concomitant immunity with age . It was determined that 86% of putatively immune individuals and 95% individuals with concomitant immunity had elevated IgG1 and IgG3 responses to Ov-103 and Ov-RAL-2 . Based on the elevated chemokine levels associated with protection in the Ov-103 or Ov-RAL-2 immunized mice , the profile of these chemokines was also analyzed in putatively immune and infected individuals; both groups contained significantly higher levels of KC , IP-10 , MCP-1 and MIP-1β in comparison to normal human sera . Moreover , human monospecific anti-Ov-103 antibodies but not anti-Ov-RAL-2 significantly inhibited the molting of third-stage larvae ( L3 ) in vitro by 46% in the presence of naïve human neutrophils , while both anti-Ov-103 and anti-Ov-RAL-2 antibodies significantly inhibited the molting by 70–80% when cultured in the presence of naive human monocytes . Interestingly , inhibition of molting by Ov-103 antibodies and monocytes was only in part dependent on contact with the cells , while inhibition of molting with Ov-RAL-2 antibodies was completely dependent on contact with the monocytes . In comparison , significant levels of parasite killing in Ov-103 and Ov-RAL-2 vaccinated mice only occurred when cells enter the parasite microenvironment . Taken together , antibodies to Ov-103 and Ov-RAL-2 and cells are required for protection in mice as well as for the development of immunity in humans . Alum-adjuvanted Ov-103 and Ov-RAL-2 vaccines have the potential of reducing infection and thus morbidity associated with onchocerciasis in humans . The development of cytophilic antibodies , that function in antibody-dependent cellular cytotoxicity , is essential for a successful prophylactic vaccine against this infection .
Onchocerca volvulus , a filarial nematode , is the etiologic agent of river blindness that infects approximately 17 million people in Africa with more than 10 million people living with skin disease and 1 million with visual impairment [1] . The current strategy for elimination of O . volvulus focuses on controlling transmission through ivermectin-based mass drug administration ( MDA ) programs . Due to factors such as the possible development of drug resistance , the need for lengthy ( >20 years ) annual drug administration , the inability to implement large-scale treatment programs in areas that are co-endemic for loiasis , it remains unlikely that onchocerciasis can be eliminated entirely through MDA with only ivermectin [2] . This realization has stimulated the search for companion intervention tools , including vaccines , to support the efforts to eliminate onchocerciasis [3–5] . A multinational consortium and initiative known as TOVA ( The Onchocerciasis Vaccine for Africa ) is working to develop a prophylactic recombinant subunit vaccine to supplement the MDA programs [3–5] . Currently , the lead candidate vaccine is comprised of two recombinant O . volvulus antigens , Ov-103 and Ov-RAL-2 [4] . Both antigens were shown to elicit protection in a mouse model of O . volvulus infection with third-stage larvae ( L3 ) [6 , 7] . Similarly , the Brugia malayi orthologous antigens were protective in a B . malayi-gerbil model , where the full life cycle of the parasite is known to develop [8] . The precise mechanisms by which protected mice and humans exposed to infection can kill larval stages of O . volvulus have not yet been fully defined . In general , it is thought that the killing of helminth parasites , which are macropathogens , is mediated by granulocytes , macrophages and antibodies using antibody-dependent cellular cytotoxicity ( ADCC ) . The Fc-receptor-bearing effector cells can recognize and kill antibody-coated parasite worms by discharging their lysosomal or granular content ( reviewed in [9–11] ) . In mice , immunization with irradiated L3 of O . volvulus induced a protective mechanism that is dependent on IgE and eosinophils [12] . Protection in mice induced by immunization with alum-adjuvanted Ov-103 , Ov-RAL-2 or Ov-103 co-administered with Ov-RAL-2 vaccines appeared to be associated with a multifactorial complex network of immune factors including specific antibodies , Th2 cytokines , chemokines , and possibly specific effector cells recruited by the chemokines such as neutrophils , monocyte/macrophages and/or eosinophils [7] . Naïve human neutrophils in the presence of polyclonal antibodies from sera samples of O . volvulus infected and putatively immune individuals have shown to be effective at killing L3 and microfilariae of O . volvulus in vitro [13 , 14] . In the gerbil-Brugia malayi infection animal model , protection induced by immunization with Bm-RAL2 and Bm-103 vaccine antigens was in part associated with the presence of antigen-specific functional antibodies that could kill B . malayi L3 in vitro in the presence of peritoneal exudate cells [8] . Notably , in both B . malayi and O . volvulus , the native proteins corresponding to Bm-RAL2 , Ov-RAL2 , Bm-103 and Ov-103 are expressed in the hypodermis of L3 and on the surface of microfilariae ( mf ) [8 , 15] , and human monospecific antibodies against Ov-103 can kill mf in vitro in the presence of neutrophils [15] . Previous studies have also shown that protective immunity in humans against L3 is associated with mixed Th1 and Th2 cytokine responses , elevated IgG1 , IgG3 and IgE cytophilic antibody responses , and possibly ADCC [16–18] . The objective of the present study was to determine whether the anti-Ov-103 and anti-Ov-RAL-2 antibody responses elicited in the vaccinated mice are essential for protective immunity . In addition , we tested whether the Ov-103 and Ov-RAL-2 antigen-specific cytophilic antibodies are also associated with protective immunity to infections with the L3 that develops in humans exposed to O . volvulus infections , i . e . in putatively immune individuals ( individuals exposed to high transmission rates of infection but had no signs or history of clinical manifestations of onchocerciasis and were negative for the presence of the O . volvulus specific 150-mer DNA repeat in skin snips over five years of surveillance ) [18] , and in infected individuals who develop concomitant immunity with age ( protection that limits newly acquired infections while adult worms and mf are not affected [17] . We also tested in vitro whether these antibodies are functional in ADCC in the presence of naïve human monocytes and neutrophils .
All the animals in this study were handled according to the National Institutes of Health ( USA ) guidelines . The animal experimentation was performed with prior approval from the Institutional Animal Care and Use Committee of Thomas Jefferson University under the protocol number 00136 . Male C57BL/6J and B6 . 129P2-Aicdatm1 ( cre ) Mnz/J ( AID-/- ) mice at 6–8 weeks of age were purchased from The Jackson Laboratory ( Bar Harbor , Maine ) . Mice were kept in the Thomas Jefferson University Laboratory Animal Sciences Facility . All mice were housed in micro-isolator boxes in a room that was pathogen free and under temperature , humidity and light cycle-controlled conditions . Mice were fed autoclavable rodent chow and given water ad libitum . The protocols used in all the human population studies were approved by the Institutional Review Board ( IRB ) of the New York Blood Center’s IRB and by the National Institutes of Health ( USA ) accredited Institutional Review Board of the Medical Research Council Kumba , Cameroon ( Kumba studies ) . Informed written consent was obtained from all adult subjects , and for children consent was obtained through both verbal assent and written consent from each subject’s legal guardian . The serum samples used for the present studies were from a repository of frozen serum samples collected during a larger clinical study performed in 1995–2000 in Kumba , an area of hyperendemicity for onchocerciasis in southwest Cameroon . The characteristics of the populations were already described in previous publications [17 , 18] . Briefly , the consenting participants enrolled in the study were born or have resided for more than 10 years in their respective villages ( Marumba I , Marumba II , Boa Bakundu , Bombanda , and Bombele ) in Cameroon . The standard skin snip test for detection of mf was performed on each subject and clinical symptoms of onchocerciasis were recorded . The average number of mf in four skin snips taken from each individual were used in estimating the individual skin mf densities . Individuals who were diagnosed clinically as infected ( presence of nodules and/or other clinical manifestations of onchocerciasis ) but were negative for skin mf were confirmed to be infected ( INF ) by establishing the presence of O . volvulus 150-mer DNA tandem repeat in the skin snips using a polymerase chain reaction ( PCR ) followed by Southern blotting using a specific internal O . volvulus probe [17 , 18] . Individuals were classified as putatively immune ( PI ) if they had no skin mf and signs of history of onchocerciasis , as well as parasitologically ( mf negative and PCR negative ) and clinically negative for O . volvulus infection during a five-year follow-up survey [17 , 18] . Notably , 75% of the PI individuals had IgG4 antibodies to Ov-16 antigen ( the antigen was provided by Dr . Thomas Nutman , NIH-NIAID ) , which confirmed that most of the PI individuals were exposed to O . volvulus infection . In comparison , 90% of the INF individuals had IgG4 antibodies to Ov-16; Ov-16 ELISA is highly specific ( >99% ) but with only 60–80% sensitivity of onchocerciasis . Ov-16 IgG4 responses are used to measure exposure to infections with L3 of O . volvulus; the antibody response can be detected 3 months and at more than 1 year before infection could otherwise be detected [19 , 20] . None of the subjects had received ivermectin treatment prior to the collection of blood . However , the consenting participants were offered and provided with ivermectin treatment by the end of their participation in the study . The procedures used to produce O . volvulus L3 were approved by an NIH accredited Institutional Review Board of the Medical Research Council Kumba , Cameroon ( Protocol 001 ) , and by the Le Comité National d’Ethique de la Recherche pour la Santé Humaine , Yaoundé , Cameroon ( Protocol 677 ) . L3 were collected from black flies ( Simulium damnosum ) that were fed on consenting infected donors . The consenting donors were offered and provided with ivermectin treatment by the end of their participation . After seven days the infected flies were dissected and the developed L3 were collected , cleaned and cryopreserved . The cryopreserved L3 were shipped to the New York Blood Center in liquid nitrogen and upon arrival in New York were stored in liquid nitrogen . All protocols using the L3 cryopreserved samples in this study were approved by the New York Blood Center’s IRB ( Protocol 321 and Protocol 603–09 ) . All L3 samples were anonymized . Informed written consent was also obtained from donors who donated blood and reside in the U . S . The protocol was approved by the New York Blood Center’s IRB ( Protocol 420 ) . The sera from these donors were treated as normal healthy controls ( NH ) . Based on previous studies , Ov-103 was expressed in PichiaPink yeast and Ov-RAL-2 was expressed in Escherichia coli . The recombinant antigens were prepared and analyzed as previously described [6] . Residual endotoxin was removed from the E . coli expressed Ov-RAL-2 using a Q anion exchange column . The level of endotoxin in the final products of both recombinant proteins was less than 20 EU/mg ( 13 . 2–19 . 3 EU/mg ) . Mice were immunized with 25 μg of the recombinant vaccine antigen formulated with 200 μg Rehydragel LV ( alum , General Chemical , Parsippany , NJ ) followed by two booster injections 14 and 28 days later . For the Ov-103 or Ov-RAL-2 alum-adjuvanted vaccines , 25 μg of each antigen in TBS was mixed 1:1 v/v with 1:5 in TBS diluted alum in a total volume of 100 μL . Mice were immunized by intramuscular injection into each caudal thigh with 50 μL of the formulated vaccine antigens . Challenge infections occurred 14 days after the final booster with 25 L3 within a diffusion chamber . Cryopreserved L3 were defrosted slowly in a two-step process , first 15 minutes on dry ice followed by a 37°C water bath . Once thawed the L3 were washed 5 times in a 1:1 mixture of NCTC-135 and Iscove’s modified Dulbecco’s medium ( Sigma , St . Louis , MO ) containing 100 U penicillin , 100 μg streptomycin ( Corning , Tewksbury , MA ) , 100 μg gentamicin and 30 μg of chloramphenicol per ml ( Millipore Sigma , St . Louis , MO ) . Diffusion chambers were constructed using 14 mm Lucite rings covered with either 5 . 0 or 0 . 1 μm pore-size Durapore membranes ( EMD Millipore , Billerica , MA ) and fused together using an adhesive containing a 1:1 mixture of 1 , 2-dichloroethane ( Fisher Scientific , Pittsburg , PA ) and acryloid resin ( Rohm and Haas , Philadelphia , PA ) . The constructed diffusion chambers were sterilized using 100% ethylene oxide followed by 12 hr aeration . The diffusion chambers were implanted in a subcutaneous pocket on the rear flank of the mice . Recovery of the diffusion chambers was performed 21 days later , and larval survival was determined based on mobility and morphology of the remaining larvae . Protective immunity was calculated based on survival of larvae within the diffusion chambers: Percent reduction = [ ( Average worm survival in control mice—Average worm survival in immunized mice ) ÷ Average worm survival in control mice] × 100 . The surviving larvae were fixed in hot 95% EtOH ( Sigma , St . Louis , MO ) and 5% Glycerol ( Fisher Scientific ) and measured using CellSens Dimension software ( Olympus , Center Valley , PA ) . Host cells within the diffusion chamber were collected and analyzed by centrifugation onto slides using a Cytospin 3 ( Shandon Inc , Pittsburgh , PA ) and then stained for differential cell counts using Hemastain 3 ( Fisher Scientific , Pittsburg , PA ) . Serum was collected at the end of the experiment for antigen specific IgM and IgG1 analysis . Maxisorp 96-well plates ( Nunc Nalgene International , Rochester , NY ) were coated with 2 μg/ml of the immunizing recombinant antigen in 50 mM Tris-HCl coating buffer pH 8 . 8 overnight 4°C . Plates were washed with deionized water between each step . Plates were blocked with borate buffer solution ( BBS ) ( 0 . 17 M boric acid , 0 . 12 M NaCl , 0 . 5% tween 20 , 0 . 025% bovine serum albumin , 1 MM EDTA , pH 8 . 2 ) at room temperature for 30 min . Individual sera were diluted to an appropriate starting concentration with BBS , serially diluted and then added to the wells . Plates were sealed and incubated at 4°C overnight . Biotinylated anti-mouse IgM and anti-mouse IgG1 ( eBioscience , San Diego , CA ) was diluted 1:250 in BBS and incubated for 1 hr at room temperature , followed by incubation with ExtrAvidin Px ( Sigma ) 1:1000 in BBS for 30 min at room temperature . One component ABTS peroxidase substrate ( KPL , Gaithersburg , MD ) was added and optical densities were read after 30 min at 405 nm in a Bio-Rad iMark Microplate reader ( Bio-Rad , Hercules , CA ) . ELISA data is presented as endpoint titers which were calculated as the serum dilution from experimental animals that had an optical density reading three times higher than the optical density recorded for control serum . One week after diffusion chamber recovery , spleens were aseptically removed from all experimental mice and homogenized into a single cell suspension . After RBC lysis , 2 x 106 cells were cultured in 96 well plates with either 10 μg of Ov-103 , Ov-RAL-2 , media or anti-CD3 mAb ( BD Biosciences , San Jose , CA ) . For a positive control , wells were pre-coated with anti-CD3 ( 0 . 5 μg/ml ) overnight at 4°C , this is the internal control to ensure the assay has worked appropriately and all T cells are generically stimulated . Each well in the assay also received 0 . 5 μL of anti-IL-4r Ab ( BD Biosciences ) . The use of 0 . 5 μL of anti-IL-4r antibody in each well is used so that the soluble IL-4 produced by cells does not immediately bind to cells and become unmeasurable in the assays . As IL-4 , IL-5 , and other TH-2 cytokines are important in the model , this is necessary to use to get quantifiable data [21–23] . Cells were incubated at 37°C for 3 days , when supernatants were collected and frozen at -20°C until use . Supernatants from the stimulated spleen cells were analyzed using Milliplex Map Kit magnetic bead panels as per the manufacturer’s directions ( Millipore Sigma ) and MAG-PIX Luminex machine ( Austin , TX ) . The results were analyzed and calculated using Milliplex Analyst software ( EMDMillipore ) . A 4-analyte cytokine panel was used to analyze the spleen cell supernatants . In this study we used 167 frozen sera samples collected from infected individuals ( 112 males and 55 females , ranging in age from 3 to 75 years ) . The infected population ( INF ) included individuals who had clinical manifestations of onchocerciasis and positive for microfilariae in their skin snip biopsies ( mf+; N = 109 , range of mf: 3–388 per skin snip ) , or those negative for skin microfilariae microscopically but positive for the O . volvulus 150-mer DNA repeat ( mf-PCR+; N = 58 ) [18] . In addition , we analyzed serum samples of 21 putatively immune ( PI ) subjects and age matched infected individuals ( n = 21 ) ranging in age from 4 to 59 years . These individuals were also assessed during enrollment for the presence of other filarial infections endemic to this region ( South West region of Cameroon ) such as Loa loa and Mansonella perstans . Both PI and INF individuals were negative for L . loa , while ~40% of the PI and ~20% of the INF were positive for M . perstans microfilariae . In our previous studies using serum samples from the same populations we have shown that protective immunity in both human populations against L3 is associated with elevated IgG1 , IgG3 and/or IgE cytophilic antibody responses against crude extracts and O . volvulus larval vaccine antigens [16–18 , 24 , 25] . Accordingly , sera obtained from INF and PI individuals were analyzed for IgG1 and IgG3 isotype antibody responses using recombinant Ov-103 and Ov-RAL-2 antigens and our established ELISA protocol [17 , 26] . Ov-103 ( 1 μg/ml ) and Ov-RAL-2 ( 1 μg/ml ) were used to coat the wells of ELISA plates , and sera samples at 1:100 dilution were added to the bound antigens ( due to the very limited volumes of sera and multiple assays that had to be performed , we were limited to single dilutions and did not perform serial dilutions to calculate endpoint titers ) . Bound antibodies were detected using a 1:1 , 000 dilution of monoclonal antibodies against IgG1 or IgG3 human subclass antibodies ( Hybridoma Reagent Laboratory , Kingsville , Md . ) . This step was followed by incubation with a 1:1 , 250 dilution of horseradish peroxidase-conjugated rabbit anti-mouse immunoglobulins ( Kierkegaard & Perry Laboratories , Inc . , Gaithersburg , Md . ) . Tetramethylbenzidine ( Sigma ) was used as the substrate for all ELISAs , and the optical density ( OD ) was read at 450 nm . Concentrations of 8 chemokines ( Eotaxin , GM-CSF , KC , IP-10 , MCP-1 , MIP-1α , MIP-1β and RANTES ) were simultaneously quantified using a MILLIPLEX MAP human cytokine/chemokine magnetic bead panel ( Millipore Sigma , Massachusetts ) in plasma from the PI ( N = 18 ) individuals as well as in INF who we classified as those with concomitant immunity ( N = 17 ) . The INF with concomitant immunity are a subset of the 109 mf+ INF individuals having high levels of mf ( mean: 144 . 5 ) , above the age of 11 ( median: 19 ) , and having high IgG3 antibodies responses to both antigens [mean: 0 . 8 ( Ov-103 ) and 0 . 9 ( Ov-RAL-2 ) ] . Plates were prepared as per the manufacturer’s directions . Chemokine concentrations were measured using MagPix Luminex100 reader . The Mean Fluorescent Intensities ( MFI ) were analyzed using MILLIPLEXTM Analyst Software ( EMD Millipore ) . The chemokine levels are expressed as pg/ml . Individuals with a positive chemokine response are defined as individuals having chemokine levels of more than the mean + three times the standard deviation ( mean+3*SD ) levels present in two pools of control normal-human sera . Monospecific human antibodies to recombinant Ov-103 and Ov-RAL-2 were purified as described by Lustigman et al . [15 , 24 , 27] . Briefly , 2 mg of recombinant Ov-103 or Ov-RAL-2 were coupled to CNBr-Sepharose 4B using the protocol provided by the manufacturer ( Pharmacia ) . Antibodies from 25 ml of pooled plasma from four infected individuals , who had high titers of anti-Ov-103 and Ov-RAL-2 antibodies , were affinity purified on the immobilized polypeptides . The mono-specificity of the eluted antibodies was confirmed by Western blot analysis . Vaccine antigen-specific negative antibodies were purified from the same pooled donor plasma using an affinity column containing an O . volvulus recombinant protein ( rOv-ASP-1 ) a putative vaccine antigen that was not able to induce protection in mice consistently [6] . The IgG endpoint titers of the anti-Ov-103 and Ov-RAL-2 purified monospecific antibodies were 1:500 , 000 and 1:1 , 000 , 000 , respectively , as determined by ELISA , while the anti-Ov-ASP-1 negative control antibodies had an anti-Ov-103 and Ov-RAL-2 cross reacting antibody titers of 1∶3 , 200 . The purified antibodies were passed through a 0 . 22 μM filter for sterilization . Human neutrophils purified from naïve PBMCs were used in the inhibition of L3 molting assays as described by Johnson EH et al . [13] . Monocytes were purified from normal human PBMCs by positive selection using anti-CD14–conjugated magnetic microbeads ( Miltenyi Biotec ) . L3 were washed in media that contains 1:1 NCTC-109 and IMDM supplemented with Glutamax ( 1x ) and 2x Antibiotic-Antimycotic ( Life Technologies ) . The larvae were diluted to 15 worms per 50 μL in serum-free media and distributed to designated wells of a 96-well plate per treatment group . 2×105 normal human neutrophils or 2×105 normal human monocytes were added to each well in 50 μL of complete medium with final concentration of 20% non-heat inactivated fetal calf serum ( Sigma ) . The anti-Ov-103 or anti-Ov-RAL-2 monospecific antibodies ( 25 μL of endpoint titer of 1:500 , 000 per well ) , or equivalent volume of anti-Ov-ASP-1 monospecific antibodies ( non-specific negative control antibodies ) were then added to each of the designated wells . Complete media without antibodies was also included as a control . Total volume per well was 200 μL . The 96-well plates were incubated at 37°C in a 5% CO2 incubator . Pooled sera from three normal healthy individuals ( NH ) and pooled sera from three individuals infected with O . volvulus who had high IgG antibody titers to both Ov-103 and Ov-RAL-2 were also used as negative and positive controls , respectively . Cultures were observed under an inverted microscope for L3 molting , the presence of the fourth-stage larvae ( L4 ) and the empty cast of the L3 , on days 6 and 12 . O . volvulus L3 molt in vitro usually by day 6 [28] . To control for possible changes in the time when molting is optimal in the presence of naive human neutrophils or monocytes , we observed the molting in control and experimental wells through day 12 . Viability of the larvae was determined on day 12 by MTT ( 3- ( 4 , 5 dimethylthiazol-2yl ) -2 , 5 diphenyl tetrazolium bromide ) staining as previously described [13] . The experiments were done in triplicates and repeated twice on separate days . Results presented are the mean ± SD of two experiments . To test whether contact between larvae , the antigen-specific monospecific antibodies and the monocytes was essential for their effect on molting , the inhibition of L3 molting assays in the presence of naïve monocytes were repeated using 96-well plates having trans-well trays with 0 . 4 μm pore size sterile polycarbonate membrane ( Sigma ) . The L3s were distributed to designated wells under the trans-well trays of the 96-well plate per treatment group in complete media as mentioned previously , while the 2×105 normal monocytes were added to each of the trans-well trays in 50 μL of complete media . The anti-Ov-103 or anti-Ov-RAL-2 purified monospecific antibodies ( 25 μL of 1:500 , 000 endpoint titer ) or the anti-Ov-ASP-1 control antibodies were added ( 25 μL ) under the trans-well trays where the L3s were added . Complete medium without antibodies was also included as a control . The 96-well plates were incubated at 37°C in a 5% CO2 incubator and the molting of L3 to L4 was monitored over the 12 days of culture . On day 12 the molting and survival was determined as described above . The experiment was done in triplicate wells and the results represent the mean ± SD of the three wells per condition . Data generated in the in vivo mouse model consisted of using 5–6 mice per group and the experiments were performed at least twice with consistent results between experiments . Data were analyzed using Systat v . 11 ( Systat Inc , Evanstown , IL ) software using multifactorial analysis of variance ANOVA with Fishers Least Significant Different test for post hoc analysis . Antibody responses and cytokine/chemokine data were analyzed by Mann-Whitney U test . Correlation of antigen specific antibody responses in infected individuals with age was performed using Spearman correlation . The in vitro inhibition of L3 molting assay data with mono-specific antibodies was analyzed by One-way ANOVA with Tukey’s multiple comparisons test , while data with NH and INF pooled sera was analyzed by Mann-Whitney test respectively using GraphPad Prism software v . 6 ( San Diego , CA ) . Probability values less than 0 . 05 were considered statistically significant .
C57BL/6J and AID-/- mice were immunized with the recombinant antigens Ov-103 or Ov-RAL-2 , with alum as the adjuvant . As expected statistically significant levels ( p < 0 . 05 ) of protective immunity were seen in C57BL/6J mice immunized with Ov-103 or Ov-RAL-2; reduction of parasite survival ranged from 35%-39% . In contrast , Ov-103 immunized AID-/- mice had no percent reduction of parasite survival , while Ov-RAL-2 immunized AID-/- mice had only 14% reduction of parasite survival , suggesting that the mice strain which lacked the ability to produce IgG , did not develop statistically significant levels of protective immunity regardless of the vaccine composition ( Fig 1 ) . Host cell recruitment to the parasite microenvironment was measured by differential cell analysis of the cells migrating into the diffusion chambers in which the worms were implanted . No differences were seen in the type or number of cells seen in diffusion chambers implanted in control and immunized mice . Nor was there a difference between cell migration into diffusion chambers implanted in C57BL/6J and AID-/- mice . In diffusion chambers recovered from C57BL/6J mice , 9 . 2 ± 9 . 1 ×104 cells were recovered in each diffusion chamber and in AID-/- mice 10 . 6 ± 10 . 3 ×104 cells were recovered . The cell composition in diffusion chambers in C57BL/6J mice was 34 ± 22% neutrophils , 2 ± 6% lymphocytes , 64 ± 23% monocytes and 1 ± 5% eosinophils , and in AID-/- mice 28 ± 18% neutrophils , 1 ± 1% lymphocytes , 71 ± 18% monocytes and 1 ± 2% eosinophils . Larvae recovered from control and immunized C57BL/6J and AID-/- mice were measured and compared to infective L3 . Worms recovered from both the control and immunized C57BL/6J and AID-/- mice had grown over the 3 weeks implanted in mice to approximately the same size with a mean of length of 655 ± 38 μm as compared to infective L3 with a mean length of 533 ± 65 μm . Antigen-specific ELISA’s were performed to measure IgM and IgG1 levels in immunized C57BL/6J and AID-/- mice . We tested only IgG1 as this was the dominant antibody subclass induced in the alum-adjuvanted vaccines [7] . Immunized C57BL/6J and AID-/- mice developed antigen-specific IgM responses , with responses in immunized AID-/- mice equivalent to those seen in C57BL/6J mice . IgG1 responses to Ov-103 and Ov-RAL-2 were elevated in C57BL/6J mice immunized with the antigens Ov-103 and Ov-RAL-2 . As was expected , there was no measurable antigen specific IgG1 response to either antigen in immunized AID-/- mice ( Table 1 ) . T cell responses in immunized mice were determined by measuring cytokine production by spleen cells recovered from immunized mice and re-stimulated ex vivo with each of the vaccine antigens . The dominant T cell responses in immunized C57BL/6J and AID-/- mice was Th2 in nature , based on elevated IL-4 and IL-5 cytokine responses to both Ov-103 and Ov-RAL-2 ( S1 Fig ) . Thus , the T cell responses elicited by both antigens in these mice were independent of the ability of the AID-/- mice to produce antibodies . To test whether humoral immune responses against the two defined vaccine antigens are also relevant to protection in humans , we tested for Ov-103 and Ov-RAL-2 antigen-specific IgG1 and IgG3 responses ( at 1:100 dilution ) in sera from O . volvulus-exposed and protected individuals classified as putatively immune , and in infected individuals who developed concomitant immunity with age . These sera samples were previously analyzed and found to contain elevated cytophilic antibodies specific to L3 as well as to other recombinant vaccine antigens such as Ov-CPI-2 , Ov-ALT-1 , and Ov-ASP-1 [17 , 24 , 26] . First , the anti-Ov-103 and anti-Ov-RAL-2 specific IgG1 and IgG3 antibody responses were determined in PI and age matched INF individuals ( Fig 2 ) . Both PI and INF individuals have high levels of IgG3 cytophilic antibodies to both Ov-103 and Ov-RAL-2 antigens ( Fig 2 ) , while the IgG1 response was significantly higher in the INF to both proteins . Notably , the number of IgG1 and IgG3 responders to Ov-103 and to Ov-RAL-2 antigens was similarly high in both PI and INF groups ( >86% IgG1 responders to both antigens; >95% IgG3 responders to both antigens ) ( Fig 2 ) . In previous studies [17] we have shown that in INF that have developed concomitant immunity with age , the IgG3 responses against crude L3 proteins as well as against Ov-ALT-1 or Ov-CPI-2 were positively correlated with age , while the IgG1 responses were elevated regardless of age [17 , 24] . Similar to the previous observations , the IgG1 responses against Ov-103 ( r = 0 . 1285 , p = 0 . 0970 ) and Ov-RAL-2 ( r = -0 . 06223 , p = 0 . 4066 ) were elevated regardless of age ( S3 Fig ) while the anti-Ov-103 and anti-Ov-RAL-2 specific IgG3 responses in the INF increased with age in response to Ov-103 ( r = 0 . 2811 , p = 0 . 0002 ) and to Ov-RAL-2 ( r = 0 . 2490 , p = 0 . 0008 ) . As this association is not very strong , even though significant , we have reanalyzed these data and compared the antibody responses in the INF individuals above 15 years of age vs . below the age of 15 . We have shown previously in the same cohort of individuals that the number of skin microfilariae plateaus in the INF by ages 10–15 years of exposure to O . volvulus [17] , which suggested that these individuals might have developed a means of limiting acquired new infections while having a stable patent infection ( microfilariae positive ) . The new analysis has shown that INF above 15 years of age had significantly higher IgG3 antibody responses to Ov-103 and Ov-RAL-2 compared to INF individuals below 15 years of age , further supporting the association of anti-Ov-103 and Ov-RAL-2 antigen-specific IgG3 antibody responses with age ( Fig 3 ) . In comparison , no significant differences were observed with the IgG1 antibody responses to the two antigens in INF below or above the age of 15 ( S2 Fig ) . These observations suggest that the elevated antibody responses to these two vaccine candidates ( Fig 3 ) are also associated with the concomitant immunity that develops in infected individuals who are exposed to multiple O . volvulus L3 infections over the years . Chemokines that recruit and activate innate cells upon infection have been shown to play an important role in the immune responses in occult O . volvulus infections [29] . Based on our previous mouse studies [7] , it appeared that chemokines associated with chemotaxis of neutrophils ( KC , MIP-1α ) , monocyte/macrophages ( MCP-1 , MIP1β ) , and eosinophils ( Eotaxin ) are also part of the multifactorial immune responses associated with protection induced by Ov-103 and Ov-RAL-2 vaccines [7] . Therefore , to explore whether this might also be reflected in the PI and INF individuals with protective immune responses to larval proteins , we analyzed the plasma levels of 8 chemokines ( Eotaxin , GM-CSF , KC , IP-10 , MCP-1 , MIP-1α , MIP-1β and RANTES ) in both groups . Notably , only the chemokine levels of KC ( neutrophils ) , MCP-1 and MIP-1β ( monocyte/macrophages ) and IP-10 , an IFN-γ-inducible protein that is a chemoattractant for monocytes and activated T cells [30 , 31] , were significantly elevated in both INF and PI individuals ( presented as % of responders ) , whereas no significant levels of the other chemokines were observed ( Table 2 ) . To test whether the elevated anti-Ov-103 and anti-Ov-RAL-2 cytophilic antibodies present in the PI and the infected individuals can function in ADCC , O . volvulus L3 were cultured in the presence of human naïve neutrophils or monocytes and monospecific human anti-Ov-103 or anti-Ov-RAL-2 antibodies . Monospecific anti-Ov-ASP-1 antibodies were used as negative control ( these antibodies did not cross-react with Ov-103 or Ov-RAL-2 ) . Molting levels were observed on day 6 and on day 12 ( S1 Table ) . The data in Fig 4 represent the L3 molting observed on day 12 . Anti-Ov-103 antibodies in the presence of neutrophils reduced significantly the molting by 46% in comparison to the control wells ( 65% of molting ) , while anti-Ov-RAL-2 antibodies did not inhibit the molting ( Fig 4A ) . The inhibition of molting ( calculated as % molting in control wells—% molting in test wells / % of molting in control wells ) in the presence of anti-Ov-103 was only 25 . 5% lower than that in the presence of anti-Ov-ASP-1 and not significant , which is potentially due to the ability of the anti-Ov-ASP-1 antibodies to also partially inhibit the molting of L3 in the presence of neutrophils ( 27 . 7% ) when compared to control wells ( Fig 4A , S1 Table ) . Pooled sera from three INF individuals completely inhibited molting in the presence of neutrophils , as compared to the 42% molting observed for worms cultured in pooled sera from normal healthy individuals ( NH ) ( Fig 4A , S1 Table ) . Both anti-Ov-103 and anti-Ov-RAL-2 antibodies , in the presence of naïve human monocytes , however , significantly inhibited the molting of larvae; 84% and 69% , respectively , when compared with control ( Fig 4B , S1 Table ) . Anti-Ov-ASP-1 antibodies also had the ability to significantly inhibit molting in the presence of monocytes ( 31 . 7% ) when compared to control wells . Nevertheless , when compared to anti-Ov-ASP-1 antibodies , both anti-Ov-103 and anti-Ov-RAL-2 antibodies in the presence of naïve human monocytes still significantly inhibited molting of larvae by 76 . 7% and 55 . 8% , respectively ( Fig 4B , S1 Table ) . Pooled sera from three INF individuals also inhibited molting completely in the presence of monocytes , as compared to 67% molting observed in pooled sera from NH individuals ( Fig 4B , S1 Table ) . None of the monospecific antibodies and cell combinations tested in this study affected significantly the survival of larvae on day 12 as compared to control wells . To test whether the inhibition of molting by the anti-Ov-103 and anti-Ov-RAL-2 antibodies in the presence of the naïve human monocytes required contact with the larvae , assays were performed using trans-wells , with the larvae placed in the bottom well and the monocytes at the top well . Anti-Ov-103 antibody-mediated inhibition of molting was found to be partially dependent on contact , with 37% inhibition of molting observed when the monocytes were separated from the larvae as compared to controls ( Fig 4C ) . When the cells were in contact with the larvae the inhibition in the presence of anti-Ov-103 was 84% on day 12 ( Fig 4B , S1 Table ) . In contrast , anti-Ov-RAL-2 antibodies did not mediate any inhibition of molting when the monocytes were not in contact with the larvae ( Fig 4C ) . C57BL/6J mice were immunized with either Ov-103 or Ov-RAL-2 and then received implanted challenge infections in diffusion chambers . The diffusion chambers were constructed with membranes of either 5 . 0 μM pore-sizes that allowed cells to enter the parasite’s microenvironment or with membranes with 0 . 1 μM pore-sizes that completely blocked cell entry . It was observed that a statistically significant level of reduction in larval survival within the diffusion chambers only occurred when cells could enter the parasite’s microenvironment; blocking cell entry into the diffusion chamber blocked the protective effects of the immune responses induced by each of the alum-adjuvanted vaccine antigens ( Fig 5 ) .
Parallel studies in mice and humans support the conclusion that protective immunity induced by the antigens Ov-103 and Ov-RAL-2 is dependent on antibodies . Based on the mouse studies , cell contact is also required for the effect on larval survival . In comparison , in the human studies cell contact ( monocytes ) is required for inhibition of molting by some antibody cell interactions ( anti-Ov-RAL-2 ) but only partially by the other ( anti-Ov-103 ) . AID-/- mice , which are unable to undergo class-switch recombination , did not develop protective immunity following immunization with alum-adjuvanted Ov-103 or Ov-RAL-2 vaccines . These mice did develop an antigen-specific IgM response but not an IgG1 response . It was shown [32] that the B cells develop normally but when there is an antigen stimulation , a much stronger IgM response develops in the AID-/- as compared to WT mice; regardless , the B cells cannot complete class switching . Apparently , IgM antibodies against Ov-103 or Ov-RAL-2 were not sufficient to kill O . volvulus larvae as compared to other nematode parasites which are also susceptible to killing through IgM dependent mechanisms [33–36] . The critical role of IgG in killing L3 within the diffusion chambers was reinforced by the observation that T cell responses in immunized AID-/- mice shared the Th2 cytokine bias seen in wild-type mice . Moreover , the number and types of cells recruited to the parasite microenvironment were equivalent to those observed in the vaccinated C57BL/6J mice . The only obvious immune defect in the immunized AID-/- mice was their IgG1 antibody response . It was therefore concluded that protection induced by vaccination of mice with alum-adjuvanted Ov-103 or Ov-RAL-2 is IgG1 antibody-dependent . Larvae recovered from control and immunized C57BL/6J and AID-/- mice were measured and had equal growth rates , indicating that there was larval killing three weeks post challenge , but no developmental regulation , in the immunized mice . Finally , studies in immunized C57BL/6J mice demonstrated that the presence of antibody alone was not sufficient to affect the survival of worms in vivo; cell contact as part of a presumptive ADCC reaction was a critical component of the protective mechanism against challenge infection . Analysis of human sera from PI and INF individuals showed that both groups had similar percentage of individuals with anti-Ov-103 and Ov-RAL-2 antigen-specific IgG1 and IgG3 responses , albeit a higher IgG1 reactivity was observed in the infected individuals . The anti-Ov-103 and anti-Ov-RAL-2 IgG3 responses in the INF individuals increased significantly with age , which is consistent with the development of concomitant immunity in these individuals over time [24] . The presence of cytophilic IgG3 antibody responses against O . volvulus larval antigens has been shown to be associated with immune protection in onchocerciasis [17 , 24 , 37 , 38] . It was postulated that one of the mechanisms through which cytophilic IgG or IgE antibodies participate in the protective immune responses in humans is by inhibiting early development of L3 to L4 , and/or by killing these parasites through ADCC [16] . Neutrophils , in the presence of sera from PI and INF were shown to be able to inhibit molting of O . volvulus larvae [13] and kill mf in vitro [39] , and neutrophils in the presence of monospecific anti-Ov-CPI-2 antibodies , inhibited molting and also killed the L3 [24] . One of the essential aspects of the ADCC mechanism is the recruitment of effector cells bearing Fc receptors , such as monocytes/macrophages , neutrophils and eosinophils that recognize and target IgG , IgA or IgE antibody coated parasitic worms by releasing their lysosomal or granular proteins [10 , 40–42] . Antibodies and ADCC were also found to be an important part of protective effector mechanisms in the O . volvulus attenuated L3 mouse model and were predicted to also function in humans living in highly endemic regions of onchocerciasis [13 , 16 , 39 , 43] . In the present study , we found the majority of both PI and INF had elevated plasma levels of the chemokines KC , IP-10 , MCP-1 and MIP-1β . KC is associated with the chemotaxis of neutrophils , while IP-10 is associated with the chemotaxis of monocytes/macrophages , dendritic cells , T cells and natural killer cells . MCP-1 and MIP-1 β are associated with the chemotaxis of monocytes [30 , 31 , 44–48] . Interestingly , in occult mf negative O . volvulus infections , serum levels of pro-inflammatory chemokines MCP-1 , MIP-1α , MIP-1β , MPIF-1 and CXCL8 were also enhanced in comparison to sera from infection-free controls [29] . Notably , in previous vaccination studies it was found that alum-adjuvanted Ov-103 vaccinated and protected C57BL/6J mice had elevated chemokines associated with activation and chemotaxis of neutrophils ( KC ) and eosinophils ( Eotaxin ) within the diffusion chambers . In comparison , MCP-1 , MIP-1β ( associated with chemotaxis of monocytes ) and MIP-1α ( associated with chemotaxis of granulocytes , particularly neutrophils ) were elevated within diffusion chambers recovered from protected mice immunized with the alum-adjuvanted Ov-RAL-2 vaccine . Because of these differences , it was proposed that the mechanism of protective immunity induced by Ov-RAL-2 might differ from that induced by Ov-103 [7] . Previous studies have shown that neutrophils purified from naïve PBMCs were able to inhibit molting and kill the L3 in the presence of sera from both the PI and the INF [13] . To test whether anti-Ov-103 and anti-Ov-RAL-2 monospecific human antibodies are functional in ADCC , we tested them first with naïve human neutrophils . Interestingly , only the anti-Ov-103 antibodies significantly inhibited the molting of O . volvulus L3 to L4 larvae ( 46% ) , while anti-Ov-RAL-2 antibody had no effect . However , both anti-Ov-103 and anti-Ov-RAL-2 antibodies significantly inhibited the molting of L3 in the presence of the naïve monocytes . This is the first study , to the best of our knowledge , in which purified monocytes were used in vitro in ADCC assays with O . volvulus L3s . None of these antibodies , however , affected the survival of the larvae under the experimental conditions tested . Moreover , the anti-Ov-103 antibody-mediated inhibition of molting was found to be partially dependent on contact with the monocytes , suggesting that soluble cytotoxic factors secreted by the monocytes may also be involved in the partial inhibition of molting . The anti-Ov-RAL-2 antibodies could not block molting when monocytes were not in direct contact with the larvae , suggesting that inhibition of molting with the anti-Ov-RAL-2 antibodies is contact-dependent . Thus , as in the mouse studies , the effect of the human Ov-103 and Ov-RAL-2 antigen-specific antibody responses might function through distinct ADCC mechanisms , even though both antigens are similarly expressed on the surface and in the glandular esophagus of L3 [8] . The soluble factors released from monocytes that cooperate with the anti-Ov-103 antibodies to partially inhibit L3 molting in vitro were not identified in this study . Soluble factors that collaborate in the nematode and trematode killing process have been identified in other systems . Strongyloides stercoralis larvae trigger the release of extracellular DNA traps by human neutrophils and macrophages , which were essential for larval killing by these cells in vitro [49] . Eosinophils release major basic protein and polymorphonuclear neutrophils ( PMN ) release myeloperoxidase that also participate in killing S . stercoralis larvae [50] . PMN cells were also shown to expel large amounts of extracellular traps that bind and form aggregates around B . malayi mf in vitro [51] . Monocytes/macrophages from rat peritoneal lavage were able to mediate ADCC against newly excysted juvenile Fasciola hepatica in vitro in a nitric oxide dependent manner [42] . Future studies will attempt to identify the soluble factors that are be involved in the inhibition of O . volvulus molting in vitro by human effector cells . Vaccine studies using the B . malayi infection model in gerbils have demonstrated that immunization with the B . malayi homologues of Ov-RAL2 or Ov-103 vaccine antigens induced protective immunity that was , at least in part , associated with the presence of antigen-specific antibodies that could kill B . malayi L3 in vitro in the presence of naïve peritoneal exudate cells [8] . The present study showed that antibodies are essential for the reduced survival of O . volvulus larvae three weeks post-challenge in mice immunized with Ov-RAL2 or Ov-103 . Neutrophils in the presence of sera from PI and INF were shown to be effective at inhibiting O . volvulus larvae molting and killing L3 and mf in vitro [13 , 14] , and in the presence of monospecific anti-Ov-CPI-2 antibodies , inhibited development and killed the L3 [24] . Human monospecific antibodies against Ov-103 also killed mf in vitro in the presence of neutrophils [15] . It was thus surprising that human monospecific antibodies to Ov-RAL2 or Ov-103 , in the presence of neutrophils or monocytes , only blocked molting but did not kill the worms ( at least by day 12 ) . A consistent observation in animal models of filarial infections is that following immunization , there are two manifestations of protective immunity . The first is inhibition of parasite development and the second is larval killing . This has been reported for Litomosoides sigmodontis [52] , Acanthocheilonema viteae [53 , 54] , B . malayi [55] and Dirofilaria immitis [56 , 57] . There are clearly differences in the immunologic characteristics between L3 and L4 of O . volvulus [28] . It has been postulated that growth retardation is a mechanism used by the immune response to keep the parasites in a stage that is more susceptible to the immune response . Therefore , from our in vitro experiments we speculate that the inhibition of molting of L3 by human monospecific antibodies and cells in vitro might be an early component of the protective immune response . Alternatively , killing of O . volvulus L3 by human monospecific antibodies to Ov-RAL-2 or Ov-103 may require more than one cell type to be present as observed with other worms [58] This might explain why there is killing of O . volvulus larvae in vivo in mice , and B . malayi L3 in vitro in the presence of naïve peritoneal exudate cells [8] yet no killing of O . volvulus L3 by human Ov-RAL2 or Ov-103 monospecific antibodies in the presence of either neutrophils or monocytes . In conclusion , this study demonstrates that the antibody responses to Ov-103 and Ov-RAL-2 antigens are integral to the development of protective immunity to O . volvulus infections in both mouse model and in humans . We have also established that this protective immunity is multi-factorial , involving a collaborative role of antibodies , chemokines and innate cells . Moreover , in vitro ADCC assays have shown that anti-Ov-103 antibodies mediated inhibition of molting of L3 with both neutrophils and monocytes , and that it was partially dependent on monocyte contact unlike with anti-Ov-RAL-2 antibodies that inhibited molting only in the presence of monocytes and this inhibition was contact-dependent . These outcomes suggest that different mechanisms of protective immunity might be induced by the two vaccine candidate antigens . Although , we did not observe larvae killing with antibodies to the individual antigens by day 12 , the observations in vitro may not represent the comprehensive range of effector mechanisms responsible for the reduction of O . volvulus larvae survival in mice and human . Regardless , it allowed us to identify some of the components associated with the immune responses against Ov-103 and Ov-RAL-2 antigens that are present in humans who develop protective immunity against O . volvulus . Future experiments in the mouse model and in vitro , including the identification of soluble factors elicited by monocytes co-cultured with mono-specific antibodies and L3 , will support a better understanding of the mechanism of protective immunity that can be induced by the alum-adjuvated Ov-103 and Ov-RAL-2 vaccines . Knowing the vaccine-induced mechanism ( s ) of protective immunity will ultimately help verify that the appropriate functional immune responses are stimulated when such vaccines are used in human clinical trials . | Onchocerca volvulus is the causative agent of river blindness that infects approximately 17 million people , mostly in Africa . The current strategy for elimination of O . volvulus focuses on controlling transmission through ivermectin-based mass drug administration programs . Due to potential ivermectin resistance , the lack of macrofilaricidal activity by ivermectin , and the prolonged time ( >20 years ) needed for successful interruption of transmission in endemic areas , additional tools are critically needed including a vaccine against onchocerciasis . Ov-103 and Ov-RAL-2 are presently the most promising vaccine candidates for a prophylactic vaccine . The mechanism of protective immunity induced in mice by the alum-adjuvanted Ov-103 or Ov-RAL-2 vaccines appear to be multifactorial with essential roles for antibodies , chemokines and the specific effector cells they recruit . In this study , we show for the first time that , anti-Ov-103 and anti-Ov-RAL-2 antibodies , chemokines and innate cells also appear to be associated with protective immunity against O . volvulus infection in humans , similar to the vaccine studies observed in the O . volvulus mouse model . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
]
| [
"blood",
"cells",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"onchocerca",
"volvulus",
"immune",
"cells",
"helminths",
"immunology",
"animals",
"onchocerca",
"physiological",
"processes",
"developmental",
"biology",
"antibodies",
"neutrophils",
"antibody",
"response",
"immune",
"system",
"proteins",
"white",
"blood",
"cells",
"molting",
"animal",
"cells",
"proteins",
"life",
"cycles",
"immune",
"response",
"biochemistry",
"eukaryota",
"cell",
"biology",
"monocytes",
"physiology",
"nematoda",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"larvae",
"organisms"
]
| 2019 | Antibody responses against the vaccine antigens Ov-103 and Ov-RAL-2 are associated with protective immunity to Onchocerca volvulus infection in both mice and humans |
The pathogenesis of dengue shock syndrome ( DSS , grade 3 and 4 ) is not yet completely understood . Several factors are reportedly associated with DSS , a more severe form of dengue infection that reportedly causes 50 times higher mortality compared to that of dengue patients without DSS . However , the results from these reports remain inconclusive . To better understand the epidemiology , clinical manifestation , and pathogenesis of DSS for development of new therapy , we systematically reviewed and performed a meta-analysis of relevant studies that reported factors in both DSS and dengue hemorrhagic fever ( DHF , grade 1 and 2 ) patients . PubMed , EMBASE , Scopus , Google Scholar , Dengue Bulletin , Cochrane Library , Virtual Health Library , and a manual search of reference lists of articles published before September 2010 were used to retrieve relevant studies . A meta-analysis using fixed- or random-effects models was used to calculate pooled odds ratios ( OR ) or event rate with corresponding 95% confidence intervals . Assessment of heterogeneity and publication bias , meta-regression analysis , subgroup analysis , sensitivity analysis , and analysis of factor-specific relationships were further performed . There were 198 studies constituting 203 data sets that met our eligibility criteria . Our meta-regression analysis showed a sustained reduction of DSS/dengue hemorrhagic fever ( DHF ) ratio over a period of 40 years in Southeast Asia , especially in Thailand . The meta-analysis revealed that age , female sex , neurological signs , nausea/vomiting , abdominal pain , gastrointestinal bleeding , hemoconcentration , ascites , pleural effusion , hypoalbuminemia , hypoproteinemia , hepatomegaly , levels of alanine transaminase and aspartate transaminase , thrombocytopenia , prothrombin time , activated partial thromboplastin time , fibrinogen level , primary/secondary infection , and dengue virus serotype-2 were significantly associated with DSS when pooling all original relevant studies . The results improve our knowledge of the pathogenesis of DSS by identifying the association between the epidemiology , clinical signs , and biomarkers involved in DSS .
Dengue infection is a major health problem in tropical and subtropical countries . Each year , more than 250 , 000 cases of DHF/DSS are reported from an estimated 50 million dengue infections [1] . Dengue disease ranges from asymptomatic or self-limiting dengue fever ( DF ) to severe dengue characterized by plasma leakage ( dengue hemorrhagic fever [DHF] , grades 1 and 2 ) that can lead to a life-threatening syndrome ( dengue shock syndrome [DSS] , grades 3 and 4 ) [2] . Recently , severe dengue was also defined by severe bleeding and/or severe organ impairment [3] . Fatal cases of dengue infection mostly occur in patients with DSS , and the mortality of DSS is reportedly 50 times higher than that of dengue patients without DSS [4] . There are no licensed vaccines or antiviral drugs against the disease , although some potential solutions are currently being studied [5] . Early appropriate treatment , vector control , and educational program are the only current methods to reduce mortality and global disease burden [3] , [6] , [7] , [8] . Therefore , the World Health Organization ( WHO ) encourages research based around markers of severity to develop new tools and methods that can reduce the healthcare burden of dengue infection in endemic countries . Several factors associated with DSS have been reported in individual studies [9] , [10] , [11]; however , the associations for some factors are not observed consistently across studies [12] , [13] , [14] , [15] . Therefore , we conducted a systematic review and meta-analysis of relevant studies to assess all reported factors associated with DSS .
Our study was performed according to the recommendations of the PRISMA statement [16] , which is available in supporting information ( Checklist S1 ) . We had developed a protocol of methods from June to August 2010 , and our protocol can be assessed in our homepage at: http://www . tm . nagasaki-u . ac . jp/hiraken/member/file/n_tien_huy/protocol_of%20systemic_review_for_dengue3 . pdf . In September 2010 , PubMed , Scopus , EMBASE , LILACS via Virtual Health Library , Google Scholar , WHO Dengue bulletin , Cochrane Library , and a manual search of reference lists of articles were searched for suitable studies . The search terms used for PubMed , EMBASE and Scopus were as follows: “dengue AND ( shock OR DSS OR severity OR severe OR “grade IV” OR “grade III” ) ” . We used “dengue” to search in LILACS and Cochrane Library . For the “Advanced Scholar Search” , we used “dengue” to fill in the field “with all of the words” , “shock OR DSS OR severity OR severe OR “grade IV” OR “grade III”” to fill in the field “with at least one of the words” , and “where my words occur” in the field “title of article” . Two independent reviewers ( NTH , TVG ) initially scanned primary titles and abstracts ( when available ) to select potential full text articles for further scrutiny according to the inclusion and exclusion criteria . The inclusion criteria were as follows: articles with reported epidemiology , clinical signs , and laboratory parameters for dengue-infected patients with shock compared with DHF . Since genetic markers are rarely reported for DSS groups [17] , [18] , we did not include these markers in this study . We used broad criteria made by original studies' authors for definition of dengue infection , DSS , and DHF to increase the number of studies included in our analysis . A subgroup analysis was used to investigate the effect of this variation on the pooled results . No restrictions were applied with respect to language , gender , patient age ( children or adult ) , or study design . Non-English reports were translated into English by authors with the help of native international students in Nagasaki University . The exclusion criteria are shown in Figure 1 . When the title and abstract were not rejected by either reviewer , the full text of the article was obtained via Nagasaki University Library and carefully reviewed for inclusion by the two reviewers ( NTH , TVG ) . Inclusion or exclusion of each study was determined by discussion and consensus between the two reviewers . When disagreement occurred , a consensus decision was made following discussion with a third reviewer ( DHDT ) . We further supplemented these searches with a manual search of articles in the WHO Dengue Bulletin , reference lists , and citation lists using the Scopus databases . For each identified factor , we performed additional factor-specific searches by adding the factor terms beside “dengue” . Data were extracted by one of two investigators ( NTH , TVG ) , and were checked by at least two of three reviewers ( NTH , TVG , DHDT ) . Disagreement was resolved via discussion and a consensus reached between the three authors . A data extraction form in an Excel file was developed by two authors ( NTH , TVG ) based on a pilot review , extraction , and calibration of 20 randomly selected studies . The data extracted included the first author , year of publication , year of patient recruitment ( the midpoint of the study's time period ) , study design ( all case or case-control ) , data collection ( prospective or retrospective ) , assignment of patients ( consecutive or random ) , country and city of origin , hospital where the patients were recruited , characteristics of the patient population ( infant , children , adult ) , criteria of dengue infection ( confirmed or clinical diagnosis ) , criteria of DSS and DHF , number of included individuals ( DSS and control DHF ) , including DF patients in the DHF group , description of blinded interpretation of factors , gender , and age at examination of included individuals . Several types of data input for factors were extracted if available and are fully described in the supplemental method ( Method S1 ) . Papers published by the same research group and studying the same factors were checked for potential duplicate data based on the year of patient recruitment and hospital where the patients were recruited . When duplications were noted , the largest data set was used for our meta-analysis . When several types of data or several methods were presented for one particular factor , we extracted all data but used the one with the least significant association ( the nearest odds ratio [OR] to one ) if that factor was significantly associated with DSS after meta-analysis . Otherwise , data with lowest and highest ORs were pooled separately to get minimal and maximal odd ratios , respectively . When data were available on different days of the disease course , values at day 4 , 3 , 5 , 2 , and 6 were favored in that order for analysis , because shock frequently occurs at day 4 and we emphasize the importance of the transition from DHF to DSS . Quality assessment was independently performed by two authors ( NTH , TVG ) . The quality of selected studies was assessed using the combined criteria suggested by Pai et al . [19] and Wells et al . [20] , because these criteria can affect the accuracy of the pooled effect size . The quality of each study included in the meta-analysis was determined across nine metrics: study design , full description of characteristic of patient population ( infant , children , adult ) , data collection ( prospective or retrospective ) , assignment of the patient ( consecutive or random ) , inclusion criteria , exclusion criteria , method quality ( description and same method for DSS and DHF groups ) , blinded interpretation of factors , and full description of dengue diagnosis . The score system is available in Table S1 . Quality assessment was also performed by discussion and consensus after the independent review of each study by two authors ( NTH , TVG ) . Meta-analyses for particular factors were performed using Comprehensive Meta-analysis software version 2·0 ( Biostat , NJ , USA ) where there was more than one study . Dichotomous and continuous variables were analyzed to compute pooled odds ratio ( OR ) and standardized mean difference , respectively when there were two groups of DSS and DHF . The standardized mean difference was then converted to OR according to the method of Borenstein et al [21] . Both dichotomous and continuous variables were combined to compute pooled odds ratio ( OR ) as previously suggested [22] to increase the number of studies included in our analysis . Therefore , the unit of continuous variables and the cut-off values of dichotomous and continuous variables were not required in the pooling result . Higher frequency of dichotomous variables and/or higher value of continuous variables in DSS compared to DHF groups resulted in a positive association of the particular variables with DSS . The event rate was pooled for the proportion of DSS among DHF/DSS cases . Only cross-sectional studies were included and DF patients in the DHF groups were excluded for this analysis of DSS prevalence . The corresponding 95% confidence intervals ( 95% CI ) of pooled effect size were also calculated using a fixed-effects or random-effects model with weighting of the studies [23] . A fixed-effects model with weighting of the studies was used when there was a lack of significant heterogeneity ( p>0 . 10 ) , while a random-effects model with weighting of the studies was used when there was heterogeneity between studies ( p≤0 . 10 ) [23] . Heterogeneity between studies was evaluated using the Q statistic and I2-test . Heterogeneity was considered statistically significant if the p-value was <0·10 [24] . I2 values>25% , 50% , or 75% were considered to represent low , moderate , or high heterogeneity , respectively [25] . To study the effect of covariates including total quality score , each parameter of quality score system , area of studies , and differences in definition of factors between studies ( category/continuous variables , diagnosis , and unit of measurement ) on the pooled effect size and the heterogeneity across studies , meta-regression analysis and subgroup analysis of a combination one or more groups were performed where there were eight or more studies assessing a particular factor [26] . The effect of covariates on the pooled effect size was considered significant when the p-value was <0·05 or its 95%CI did not overlap with the original one . To evaluate the presence of publication bias , we performed Begg's funnel plot [27] and Egger's regression test [28] , [29] when there were five or more studies assessing the association of a particular factors with DSS . Publication bias was considered significant when the p value was <0·1 . If publication bias was found , the trim and fill method of Duvall and Tweedie was performed by adding studies that appeared to be missing [30] , [31] to enhance the symmetry [26] . The adjusted pooled effect size and its 95% CI were computed after the addition of potential missing studies . We further performed a sensitivity analysis by removing each study from the meta-analysis to investigate the effect of each study on the association . Cumulative meta-analysis was also carried out to test the effect of a few of the largest or smaller studies on the effect size by repeatedly performing a meta-analysis each time a new study was added according to its sample size ( reversed variance of logOR or log of rate of event ) . The p-value for multiple comparisons was not adjusted because it may increase the likelihood of type II errors [32] , [33] . Instead , to reduce the false discovery rate , a confidence interval , meta-regression , subgroup analysis , sensitivity analysis , and interpretation of across studies were conducted to give complement information to p-value . Thus , statistical significance was defined as p-value was <0 . 05 ( two-tailed test ) or its 95%CI did not overlap with the original one . Analysis of factor-specific relationships with DSS was also performed when there were three or more categories reported for a particular factor using the GraphPad Prism 5 ( GraphPad Software , Inc . , San Diego , CA , USA ) . This analysis was performed according to the dose-response relationship as previously reported [34] , [35] . Briefly , the midpoint of each category for a particular factor was assigned to plot against the natural logarithm of OR ( logOR ) or rate of event . If no upper bound was available , we assumed it to be the same amplitude as the preceding category . When no lower bound was reported , we assigned it a value of zero . We used a mixed-models analysis [36] to test a potential nonlinear factor-specific relationship between factor and DSS by using polynomial , sine wave , and exponential regression models weighing on the sample study size . When the candidate models were nested , we used likelihood ratio tests ( F-test ) to examine whether the more complex model was a better fit . When comparing two non-nested models , we used the Akaike information criterion [37] , which indicates twice the number of parameters of the model minus twice the maximized log-likelihood . The model with lowest Akaike information criterion value was chosen for fitting .
The initial screening of the databases for title and abstract yielded 5612 papers , of which 798 papers were chosen for full text reading . A total of 600 articles were excluded for one of the reasons listed in Figure 1 . Finally , 198 studies were selected for final analysis with agreement between the two reviewers at 93% ( Cohen's kappa = 0·810 ) . Three selected studies separately reported data into two data sets of infants , children , and/or adult groups [38] , [39] , [40] , while one study divided data into three data sets of infants , children , and adult groups [41]; hence a total of 203 data sets were included in the final meta-analysis . Characteristics of the included studies are outlined in Table S2 . Most studies were performed in Asia ( n = 182 ) ; only 12 , four , and five studies were from the Caribbean , South American , and French Polynesia , respectively . More studies were prospective ( n = 150 ) than retrospective or not mentioned ( n = 53 ) . A total of 88 studies were case-control assessments and 115 were cross-sectional studies . Eight studies included infants , 83 studies enrolled children , 22 studies recruited adults , 54 studies enrolled both infants and children , 18 studies reported both children and adults , 14 studies included all types of subject , and four studies did not provide this information . Clinical diagnosis was used as for dengue infection definition in 14 studies , while serology , PCR , and virus isolation were used for confirmation of all dengue-infected patients in 170 studies . Fourteen studies did not report the criteria for dengue infection . The classification of DSS and DHF varied across studies , but most studies used the WHO 1997 criteria ( n = 168 ) , 28 studies used the Nimmannitya criteria [42] , while the other seven studies simply classified the diseases as shock versus non-shock group ( Table S2 ) . In terms of the quality of included studies , agreement between the two reviewers was 91% ( Cohen's kappa = 0·808 ) . Two studies scored the maximal points ( 9 ) ; the range of total points of included studies was two to nine ( Table S2 ) . A total of 242 factors were reported in at least one study . More than half of them ( n = 130 ) were available in only one study and then were not assessed by meta-analysis , but the association with DSS derived from the original study is shown in the Table S3 . There were 112 factors that were reported in two or more studies , and the results of our meta-analysis including the references of included studies for each factor are shown in Table S4 . Among 112 factors , 72 factors were investigated in less than eight studies and were not interpreted here because drawing a conclusion is limited when there is very few included studies [26] . There is no clear cut off point for the minimal number of studies included in a meta-analysis to draw a conclusion . Cox et al suggest a description of individual studies is better than a meta-analysis [43] . Several studies chose eight as a cut off number for studies included in a meta-analysis to perform a regression analysis [44] and assessment of publication bias [45] . Finally , 40 factors were fully analyzed and interpreted when there were eight or more studies assessing the particular factor . Of these interpreted factors , 23 factors were found to be significantly associated with DSS ( Table 1 ) . In 80 published studies of cross-sectional design , our pooled results showed that the proportion of DSS among DHF/DSS was 28·5% ( 95% CI: 24·7 - 32·6 ) with high heterogeneity ( p-value for heterogeneity <0·001 , I2 = 95 ) . Sub-analysis further demonstrated that DSS prevalence in adults ( 17·7%; 95% CI: 10·1 - 29·4; in 11 studies recruited only adults ) was significantly lower than in children ( 37·4%; 95% CI: 29·6 - 45·9; in 26 studies recruited only children ) . Meta-regression analysis showed a trend in the proportion of DSS among DHF/DSS cases that gradually decreased over a period of 40 years , but the trend was not statically significant ( p = 0·089 , Figure 2A ) . However , after excluding three studies in South America , the decreased trend became significant ( p = 0·040 , Figure 2B ) . The decreased trend was significant for 49 studies in Southeast Asia ( p = 0·045 , Figure 2C ) and particularly steep for 23 studies in Thailand ( p = 0·004 , Figure 2F ) . The reduced trend was also observed for 16 studies in South Asia but was not statistically significant ( p = 0·6 , Figure 2D ) , probably due to the small number and shorter duration of studies . The proportion of DSS among DHF/DSS was low in Caribbean countries ( pooled prevalence: 19·7%; 95% CI: 15·1 - 25·4; n = 7 ) , explaining the non-significant reduction in prevalence over a period of 25 years ( p = 0·5 , Figure 2E ) . Other covariates including quality score , sample size , area , country , study design , data collection , different inclusion criteria for DHF , different inclusion criteria for DSS , assignment of the patient ( consecutive or random ) , blinded interpretation of factors , and confirmation of dengue diagnosis had no effect on the pooled prevalence and heterogeneity across studies in Southeast Asia and Thailand , separately . No evidence of publication bias was found using Begg's funnel plot [27] and Egger's regression test [28] , [29] . Meta-analysis of 37 studies for gender difference showed a significant association with DSS ( OR: 1·37 , 95% CI: 1·17-1·60 ) . Removing any study among selected studies had little effect on the pooled OR ( Table 1 ) . Cumulative meta-analysis by repeated meta-analyses each time a new study was added according to sample size demonstrated that this significant association was established without the 17 largest studies . This finding suggested a strong association between female gender and DSS . Significant heterogeneity was found among studies of females; however , removing three studies in the Caribbean area lowered the heterogeneity degree ( p value for heterogeneity = 0·075 , I2 = 27 ) . Further , removing one study from Colombia made the data homologous ( p value for heterogeneity = 0·489 , I2 = = 0 ) , but did not significantly affect the summary effect size . Meta-regression and sub-analysis for several co-variables including quality of study , year of publication , year of patient recruitment , area/country of the study , study design ( all case or case-control ) , study that included DF patients in the DHF group , data collection ( prospective or retrospective ) , assignment of the patient ( consecutive or random ) , confirmed diagnosis of dengue , different criteria of DSS and DHF , and characteristic of patient population ( infant , children , adult ) were performed to evaluate the effect of these co-variables on the summarized effect size and heterogeneity ( Table S5 ) . The homogeneity was present when pooling 16 studies in children with a positive correlation between female and DSS ( OR: 1·23; 95% CI: 1·03-1·51 ) and five studies in adults with a significant association between female and DSS ( OR: 1·32 , 95% CI: 0·94-1·87 ) . Moreover , subgroup analysis of 24 prospective studies showed an identical pooled OR 1·36 ( 95% CI: 1·17-1·59 ) with a homogenous characteristic ( p value for heterogeneity = 0·175 , I2 = 21 ) . Other co-variables including quality of study , year of publication , year of patient recruitment , area/country of the study , study design ( all case or case-control ) , study that included DF patients in the DHF group , assignment of the patient ( consecutive or random ) , confirmed diagnosis of dengue , and different criteria of DSS and DHF did not affect the summarized effect size and the heterogeneity . No evidence of publication bias was found for female gender as a factor ( Table 1 ) . Pooled odds ratio showed that age was negatively associated with DSS ( OR: 0·50 , 95% CI: 0·36 - 0·70 ) . However , pooling all studies gave high heterogeneity and publication bias ( p<0·001 ) , probably due to large variation of population age in the studies ( children/adults ) . Adding 13 missing studies on the left to enhance the symmetry using the trim and fill method of Duvall and Tweedie ( random effect ) gave a stronger association with DSS ( adjusted OR: 0·27 , 95% CI: 0·17-0·42 ) . Because only five and two studies investigated adults and infants , respectively , we could not analyze the age factor in these sub-groups . Pooling 26 studies of children gave a negative association with DSS ( OR: 0·67 , 95% CI: 0·54 - 0·84 ) . We further investigated the average age of children in DSS and DHF groups in South East Asia . The difference in average age was slightly wider over a period of 40 years , but the trend was not statistically significant ( p = 0·37 , Figure 3A ) . There was strong evidence of increasing average age of children in both DSS and DHF groups in this area ( p<0·05 , Figure 3B–C ) , agreeing with a previous study [46] . The age increase of DHF children ( slope = 0·084 ) was slightly faster than that of DSS ( slope = 0·067 ) . Upon pooling nine studies , malnutrition was positively associated with DSS ( OR: 1·19 , 95% CI: 1·00-1·41 , Figure 4A ) . No evidence of heterogeneity was found for malnutrition ( p = 0·37 , I2 = 8 ) , but publication bias was observed by an asymmetric funnel plot ( figure 4B ) and Egger's test ( p = 0·03 , Table 1 ) . The definition of malnutrition differed between studies: three studies did not provide definitions [47] , [48] , [49]; one study used weight-for-height [50] , while other five [15] , [51] , [52] , [53] , [54] used weight-for-age to assess this factor . Removing any subgroups of no definition and weight-for-height did not affect the association . Furthermore , sub-analysis of five studies using the weight-for-age also gave a positive association with DSS ( OR: 1·29 , 95% CI: 1·05-1·58 ) without any evidence of publication bias . However , removal of the largest study [15] eliminated the association of malnutrition and DSS ( OR: 0·97 , 95% CI: 0·76-1·25 , Figure 4C ) and publication bias ( p = 0·12 , Figure 4D ) . Normal nutrition was inversely linked with DSS in nine studies ( OR: 0·87 , 95% CI: 0·77-0·99 ) without evidence of heterogeneity ( p = 0·26 , I2 = 21 ) or publication bias ( p = 0·43 ) . No report significantly demonstrated a positive association of this factor with DSS , while one study showed a negative association with DSS [50] . Two studies did not give definition of normal nutrition [47] , [49]; two study used weight-for-height [50] , [55] , while other five [15] , [51] , [52] , [53] , [54] used weight-for-age to assess this factor . Removing the subgroup of no definition [47] , [49] gave a significant association with DSS ( OR: 0·86 , 95% CI: 0·76-0·98 ) with very low heterogeneity ( p = 0·42 , I2 = 0 ) and without publication bias ( p = 0·83 ) . However , thought sub-analysis of five studies using the weight-for-age still gave an OR <1 but the statistical significant association was lost ( OR: 0·92 , 95% CI: 0·80-1·05 ) . Furthermore , a sensitivity analysis showed that removing any of three studies of Junia et al [50] , Kalayanarooj et al [15] , or Pham et al [55] resulted in a loss of statistical association but the ORs were less than one ( 0·05<p<0·2 ) . Obesity/overweight was not associated with DSS in eight studies ( OR: 1·31 , 95% CI: 0·91-1·88 ) . No evidence of heterogeneity or publication bias was found for this factor . One study assessed the obesity/overweight using weight-for-height [50] , two studies did not defined the criteria of obesity/overweight [47] , [49] , while other five studies assessed this factor using weight-for-age . Removing any or all studies by Junia et al [50] , Widagdo et al [47] , and Basuki et al [49] did not affect the pooled result . No effect of quality score and sample size of included studies on pooled results were observed for three nutritional factors . Neurological signs were defined as any sign of restlessness , irritability , dizziness , drowsiness , stupor , coma , or convulsion . Meta-analysis of 15 studies gave a high pooled OR of 4·66 ( 95% CI: 1·70-12·8 ) with high degree of heterogeneity . Subgroup analysis of five studies in Thailand revealed a relative stronger correlation with DSS ( OR: 12·7 , 95% CI: 6·67-24·3 ) without evidence of heterogeneity ( p = 0·42 , I2 = 0 ) . Similarly , summary of seven case-control studies showed a slightly stronger association of encephalopathy with DSS ( OR: 5·31 , 95% CI: 2·68-10·5 ) without evidence of heterogeneity ( p = 0·25 , I2 = 24 ) . Sub-group analysis of nine studies with consecutive or random enrolment showed high heterogeneity ( p<0·001 , I2 = 89 ) and non-association with DSS ( OR: 3·13 , 95% CI: 0·73-13·4 ) . However , removing the study of Kamath et al [56] lowered the heterogeneity ( p<0·004 , I2 = 67 ) and made the factor strongly associated with DSS ( OR: 5·39 , 95% CI: 2·13-13·7 ) . Other co-variables ( quality of study , year of publication , year of patient recruitment , study that included DF patients in the DHF group , confirmed diagnosis of dengue , and different criteria of DSS and DHF ) did not affect the summarized effect size or heterogeneity . No evidence of publication bias was found for this factor . Among digestive factors , vomiting/nausea and abdominal pain were identified as associated factors for DSS in 14 ( OR: 1·43 , 95% CI: 1·15-1·78 ) and 17 ( OR: 2·26 , 95% CI: 1·76-2·89 ) studies , respectively . No evidence of heterogeneity was found for vomiting/nausea , while clear evidence of heterogeneity ( p = 0·014 , I2 = 48 ) was found for abdominal pain . Cumulative analysis of abdominal pain showed that the eight smallest studies were enough to get a significant association with DSS . Removal of one outlier study [57] eliminated the heterogeneity across studies ( p = 0·15 , I2 = 27 ) , but did not significantly affect the pooled result ( OR: 1·92 , 95% CI: 1·72-2·14 ) . Sub-group analysis of abdominal pain studies that used WHO criteria for DSS and DHF also provided a homogenous result , probably due to loss of the outlier study [57] . Meta-regression analysis revealed an increased association with DSS according to publication year ( p = 0·049 ) and quality score ( p = 0·004 ) , suggesting a possible under-estimated OR when pooling low-quality and early studies for abdominal pain . Gastrointestinal bleeding was identified as a positive associated factor for DSS ( pooled OR: 1·84 , 95% CI: 1·42-2·39 ) , while positive tourniquet test , skin hemorrhages , petechiae , hematuria , and hemoptysis were not associated after pooling more than 10 studies ( Table S4 ) . No evidence of heterogeneity was found in 18 studies of gastrointestinal bleeding . No associations between DSS and ecchymoses/purpura and gum or nose bleeding were found in eight and nine studies , respectively , further suggesting that skin and mucosal bleeding were not associated with DSS . All plasma leakage signs ( hemoconcentration , pleural effusion , ascites , hypoalbuminemia , and hypoproteinemia ) , with more than nine studies for each factor , were strongly associated with DSS after pooling relevant studies ( Table 1 ) . Significant heterogeneity among sub-group analysis for studies of pleural effusion , ascites , and hepatomegaly remained , and no effect of co-variables on the pooled effect size was found on those factors . Only subgroup analysis for hemoconcentration in South Asia revealed homogeneity across 10 studies without significant change of effect size ( OR: 1·77 , 95% CI: 1·27 , 2·46 ) . After removal of two outliers [58] , [59] from 13 primary studies , the summary result of hypoalbuminemia became homogenous without significant changes in the OR . Meta-regression analysis showed that early or low-quality studies significantly overestimated the pooled OR for hypoalbuminemia; however , the summary effect size was not significantly changed after removing the four earliest studies or lowest quality studies . Evidence of publication bias was present for pleural effusion using Egger's test ( p = 0·007 , Table 1 ) . Adding four missing studies to enhance the symmetry using the trim and fill method of Duvall and Tweedie ( random effect ) gave a stronger association with DSS ( original pooled OR: 10·4 , 95% CI: 5·47-19·6; adjusted OR: 15·8 , 95% CI: 7·95-31·6 ) . We further plotted the logOR against hematocrit ( Hct ) categories . The result showed that a first order polynomial model ( straight line ) was the best fit compared to the second order polynomial , sine wave , and exponential regression models . The fitted trend revealed an increase of logOR at 20·5% ( 95% CI: 4·65-36·36 ) for every 1% increment of Hct ( Figure 5A ) . Pooling 26 studies for both alanine transaminase ( ALT ) and aspartate transaminase ( AST ) resulted in summary ORs of 2·15 ( 95%CI: 1·47-3·15 , p<0·001 ) and 2·08 ( 95%CI: 1·39-3·12 , p<0·001 ) , respectively . Cumulative analysis showed that removing at least 11 and 7 of the largest studies was needed to change ALT and AST , respectively , into factors not associated with DSS , indicating a strong association between those two parameters and DSS . There were no effects of co-variables on the pooled OR and heterogeneity for ALT . A subgroup analysis for AST across nine studies in South Asia showed homogeneity but an identical effect size ( OR: 2·19 , 95% CI: 1·33-3·61 ) . There were a decreasing trend of OR for AST over the recruitment year and publication year of studies; however , after removal of two earliest studies [60] , [61] , the effect of recruitment year and publication year on OR was lost , suggesting the two early studies overestimated the association with DSS . Hepatomegaly was strongly associated with DSS after pooling 28 relevant studies ( Table 1 ) . Significant heterogeneity among sub-groups remained , and no effect of co-variables on pooled effect size was found . No evidence of publication bias was found . A negative association of blood platelet count was demonstrated by pooling 47 studies . The relationship between DSS and blood platelet count is shown in the ( Figure 5B ) . Based on the F-test and the Akaike information criterion values [37] , the relationship between the logOR and blood platelet count was best interpreted by a linear equation with a raise of logOR at 33·2% ( 95% CI: 20·3-46·1 ) for every 10 , 000 platelet cells decrement . White blood cells and leukopenia were not associated with DSS in the meta-analysis of 15 and nine studies , respectively ( Table S4 ) . Meta-analysis of coagulators showed a positive association between DSS and prolonged prothrombin time ( PT ) and activated partial thromboplastin time ( APTT ) . One study [62] only reported risk ratio and could not be combined with other the 15 studies for both factors . However , this study also suggested a positive association with DSS , agreeing with the pooling effect size of the other 15 studies . Heterogeneity disappeared after removing two outliers regarding PT factor [63] , [64] or computing values only in the infant/children group , but the summary OR was unchanged . Subgroup analysis for APTT factor showed homogenous results in children's group . Evidence of publication bias was present using Egger's test ( p = 0·017 , Table 1 ) . Adding one missing study to enhance the symmetry using the trim and fill method of Duvall and Tweedie gave a similar result ( adjusted OR: 5·18 , 95% CI: 2·19-12·2 ) . Inverse associations between DSS and fibrinogen level were found in nine studies ( Table 1 ) . Primary infection was reported as an inverse associated factor in two of 37 studies ( Table 1 ) . However , most of the ORs were consistently less than one , and pooling of 37 studies gave a strong negative association with DSS ( OR: 0·47 , 95% CI: 0·37- 0·60 , p<0·0001 ) without evidence of heterogeneity . Cumulative analysis by repeatedly pooling each time after adding a new study according to sample size showed that an inversely significant association was established without the 28 largest studies , further supporting a strong inverse association between primary infection and DSS . Meta-regression analysis revealed a significant decreased OR of DSS in larger studies ( p-value for slope = 0·036 ) , and no effect of other co-variables on the summary effect . Meta-analysis of 40 studies for secondary infection gave a positive association with DSS with high degree of heterogeneity ( Table 1 ) . In the sensitivity analysis , the summary OR was not significantly affected by any co-variable and when excluding any study or the six largest studies . Dengue virus serotype 2 ( DENV-2 ) was found to be an associated factor for DSS ( OR: 1·66 , 95% CI: 1·09 - 2·55 ) , whereas DENV-1 , DENV-3 , and DENV-4 were not significantly associated with DSS after pooling the whole population . Significant heterogeneity among studies of these serotypes was found . However , subgroup analysis of eight DENV-2 studies in Thailand showed a homogenous result with a slightly higher association with DSS ( OR: 1·99 , 95% CI: 1·63 - 2·43 ) , but no association between DSS and DENV-2 in other individual countries or in all countries after excluding Thailand ( Figure 6 ) .
Our meta-analysis indicated a sustained reduction of DSS/DHF ratio in Southeast Asia , suggesting an improvement in clinical recognition and management over the 40-year period ( Figure 3 ) . This improvement may have been due to educational programs and early treatment of rehydration [65] . The upward shift of average age in South East Asia ( Figure 3 ) and previous study [46] ) , coupled with the lower risk of DSS in older age [4] , [66] could be another reason for the reduction of DSS/DHF ratio over this period . It is well known that young children have an increased risk of severe dengue infection [4] , [55] , probably due to a combination of timing of secondary infection , development of protective immune responses after infection with four strains [67] , and increased microvascular fragility in younger children [68] . Our pooled result of 37 studies further strengthened the negative association of age with DSS . It is thought that normal nutrition is a risk factor of DSS , while malnutrition is a protective factor due to suppressed immune activation in malnourished children [12] , [13] , [14] . However , both factors had conflicting effects on DSS risk in previous studies [15] , and our pooled results of all relevant studies suggested that malnutrition was a positive associated factor for DSS , while normal nutrition helped protect against DSS . However , these associations should be interpreted with caution , because removal of one particular study led to a loss of statistical association of DSS with both factors . Thus , more well-designed prospective studies using both weight-for-height and weight-for-age for nutritional status [69] are required to confirm these results . The association between being female gender and risk of DSS is not fully understood , but may be explained by gender differences in seeking healthcare as well as physical characteristics [4] . Several manifestations of neurological disorders and liver damage were found to be strongly associated with DSS in our meta-analysis , probably due to the sequelae of shock , systemic inflammation , and direct viral invasion into the organs [70] . As expected , DSS was associated with signs of plasma leakage including hemoconcentration , pleural effusion , ascites , hypoalbuminemia , and hypoproteinemia to varying degrees . Based on the curve fitting for the included studies , the logarithm scale for the OR was raised approximately 20% for every 1% increment of Hct . Notably , only gastrointestinal bleeding was found to be associated with DSS , while other cutaneous and mucosal bleedings were not . Thrombocytopenia is a well-known marker of dengue severity . The WHO recently suggested that a “rapid drop in platelet count” is a warning sign for dengue severity [3] . Using a value-specific analysis , we demonstrated that the logarithm scale of OR for DSS over DHF was increased 33% for every 10 , 000 platelet cells decrement . More studies on the association of kinetic platelet count with the risk of dengue severity are recommended to define the “rapid drop in platelet count” and predict the patients who will develop the severity . Our results further confirm a positive association of DSS with secondary dengue infections , probably related to the role of antibody-dependent enhancement in DSS pathogenesis . We found that DENV-2 was associated with DSS in Thailand , but not in other countries . This could be due to differences in primary/secondary infections , a highly virulent Asian DENV-2 genotype circulating in Thailand [71] , ethnic factors , or the small number studies in other countries . Future studies need to analyze the risk of DSS in separate primary or secondary infected patients and simultaneously detect four DENV strains . One limitation of our study is that the effect of primary/secondary infection , data during epidemics , other underlying diseases , and early treatment could not be assessed because of limited information provided in the included studies . Secondly , the variability in study designs , diagnoses , population selection , and severity definitions [72] could limit the interpretation of our results . In addition , several factors , particularly cytokines and lipid profiles , were found to be associated with DSS , but the number of studies was small [11] . Thus , the results for these factors must be interpreted with caution and more studies are required to validate them . Thirdly , we transformed standardized mean difference to OR based on the assumption that the continuous data had a logistic distribution [21] . Though the method is reportedly reasonable [73] , some continuous could have skewed distribution . However , our sensitivity analyses showed that separately pooling studies with category and continuous variables , respectively , did not affect the summarized effect size and the heterogeneity ( Table S5 ) . Fourthly , the results of Figure 5 should be interpreted with caution because of ‘ecological fallacy’ , which happen when the relationship between risks of DSS with particular factors across studies could be different from that relationship within studies [26] . Another limitation of the study is that we could not investigate the factors in a time-course profile in which a factor may change quickly before and after defervescence , potentially leading to high heterogeneity across studies . In conclusion , although some factors were found to have contradictory effects , this meta-analysis identified significant associations between 21 factors ( including age , female sex , neurological signs , nausea/vomiting , abdominal pain , gastrointestinal bleeding , hemoconcentration , ascites , pleural effusion , hypoalbuminemia , hypoproteinemia , hepatomegaly , levels of alanine transaminase and aspartate transaminase , thrombocytopenia , prothrombin time , activated partial thromboplastin time , fibrinogen level , primary/secondary infection , and DENV-2 ) and DSS . These results may improve the understanding of the epidemiology , clinical manifestation , and pathogenesis of DSS . | Dengue is one of the most common viral diseases transmitted by infected mosquitoes . It may range from asymptomatic or self-limiting dengue fever ( DF ) to severe dengue characterized by plasma leakage ( dengue hemorrhagic fever , DHF ) and dengue shock syndrome ( DSS ) . Death from dengue infection occurs mostly in DSS , and the mortality of DSS is reportedly 50 times higher compared to that of dengue patients without DSS . Several factors associated with DSS have been reported in individual studies; however , the associations for some factors are not observed consistently across studies . Therefore , we conducted a systematic review of the literature to clarify this issue . The study showed persons with younger age , female sex , neurological signs , nausea/vomiting , abdominal pain , gastrointestinal bleeding , increased hemoconcentration , ascites , pleural effusion , hypoalbuminemia , hypoproteinemia , hepatomegaly , increased level of ALT or AST , thrombocytopenia , coagulation dysregulation , secondary infection , and infection of dengue virus serotype 2 are more likely to have DSS . This result improves our knowledge of the clinical manifestation and pathogenesis of DSS . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
]
| []
| 2013 | Factors Associated with Dengue Shock Syndrome: A Systematic Review and Meta-Analysis |
A major challenge in analyzing animal behavior is to discover some underlying simplicity in complex motor actions . Here , we show that the space of shapes adopted by the nematode Caenorhabditis elegans is low dimensional , with just four dimensions accounting for 95% of the shape variance . These dimensions provide a quantitative description of worm behavior , and we partially reconstruct “equations of motion” for the dynamics in this space . These dynamics have multiple attractors , and we find that the worm visits these in a rapid and almost completely deterministic response to weak thermal stimuli . Stimulus-dependent correlations among the different modes suggest that one can generate more reliable behaviors by synchronizing stimuli to the state of the worm in shape space . We confirm this prediction , effectively “steering” the worm in real time .
The study of animal behavior is rooted in two divergent traditions . One approach creates well-controlled situations , in which animals are forced to choose among a small discrete set of behaviors , as in psychophysical experiments [1] . The other , taken by ethologists [2] , describes the richness of the behaviors seen in more natural contexts . One might hope that simpler organisms provide model systems in which the tension between these approaches can be resolved , leading to a fully quantitative description of complex , naturalistic behavior . Here we explore the motor behavior of the nematode , Caenorhabditis elegans , moving freely on an agar plate [3]–[9] . Though lacking the full richness of a natural environment , this unconstrained motion allows for complex patterns of spontaneous motor behaviors [10] , which are modulated in response to chemical , thermal and mechanical stimuli [11]–[13] . Using video microscopy of the worm's movements , we find a low dimensional but essentially complete description of the macroscopic motor behavior . Within this low dimensional space we reconstruct equations of motion which reveal multiple attractors—candidates for a rigorous definition of behavioral states . We show that these states are visited as part of a surprisingly reproducible response of C . elegans to small temperature changes . Correlations among fluctuations along the different behavioral dimensions suggest that some of the randomness in the behavioral responses could be removed if sensory stimuli are delivered only when the worm is at a well defined initial state . We present experimental evidence in favor of this idea , showing that worms can be “steered” in real time by appropriately synchronized stimuli .
We use tracking microscopy with high spatial and temporal resolution to extract the two-dimensional shape of individual C . elegans from images of freely moving worms over long periods of time ( Figure 1A; see Materials and Methods ) . Variations in the thickness of the worm are small , so we describe the shape by a curve that passes through the center of the body ( Figure 1B ) . We measure position along this curve ( arc length ) by the variable s , normalized so that s = 0 is the head and s = 1 is the tail . The position of the body element at s is denoted by x ( s ) , and we sample this function at N = 100 equally spaced points along the body . These variables provide an essentially complete description of the motor output . We analyze the worm's shapes in a way intrinsic to its own behavior , not to our arbitrary choice of coordinates ( Figure 1 ) . The intrinsic geometry of a curve in the plane is defined by the Frenet equations [14] , [15] , ( 1 ) ( 2 ) where ( s ) is the unit tangent vector to the curve , ( s ) is the unit normal to the curve , and κ ( s ) is the scalar curvature . If the tangent vector points in a direction κ ( s ) , then κ ( s ) = dθ ( s ) /ds . Curvature as a function of arc length , θ ( s ) , thus provides a “worm–centered” description , but in practice this involves taking two derivatives and thus is noisy . As an alternative , we describe the curve by θ ( s ) , but remove the dependence on our choice of coordinates by rotating each image so that the mean value of θ along the body always is zero; this rotated version of θ ( s ) contains exactly the same information as κ ( s ) . Although the worm has no discrete joints , we expect that the combination of elasticity in the worm's body wall and a limited number of muscles will lead to a limited effective dimensionality of the shape and motion . In the simplest case , the relevant low dimensional space will be a Euclidean projection of the original high dimensional space . If this is true , then the covariance matrix of angles , C ( s , s′ ) = 〈 ( θ ( s ) –〈θ〉 ) ( θ ( s′ ) –〈θ〉 ) 〉 will have only a small number of nonzero eigenvalues . Figure 2A shows the covariance matrix , and its smooth structure is a strong hint that there will be only a small number of significant eigenvalues; this is shown explicitly in Figure 2B . Quantitatively , over 95% of the total variance in angle along the body is accounted for by just four eigenvalues . Note that the contribution of the variance is inhomogeneous along the body curve . For example the fourth eigenworm makes a small contribution to the variance overall , but captures a large percentage of the variance within 5% of the head and tail region ( Figure 2D ) . Associated with each of the eigenvalues λμ is an eigenvector uμ ( s ) , sometimes referred to as a ‘principal component’ of the function θ ( s ) . If only K = 4 eigenvalues are significant , then we can write the shape of the worm as a superposition of ‘eigenworm’ shapes , ( 3 ) where the four variables {aμ} are the amplitudes of motion along the different principal components , . We see in Figure 2C that these modes are highly reproducible from individual to individual . Thus far we have considered only worms moving in the absence of deliberate sensory stimuli . Do the worms continue to move in just a four dimensional shape space when they respond to strong inputs ? To test this , we delivered intense pulses of heat ( see Materials and Methods ) , which are known to trigger escape responses [16] . We see in Figure 2D that we still account for ≈95% of the shape variance using just four modes , even though the distribution of shapes during the thermal response is very different from that seen in spontaneous crawling . We conclude that our four eigenworms provide an effective , low dimensional coordinate system within which to describe C . elegans motor behavior . The projection of worm shapes onto the low-dimensional space of eigenworms provides a new and quantitative foundation for the classical , qualitative descriptions of C . elegans behavior [10] . The first two modes are sinuous ( although not exactly sinusoidal ) oscillations of the body shape ( Figure 2C ) ; they form a quadrature pair , so that different mixtures of the two modes correspond to different phases of a wave along the body . Indeed , the probability distribution of the mode amplitudes , ρ ( a1 , a2 ) , shows a ring of nearly constant amplitude ( Figure 3A ) . Sampling images around this ring reveals a traveling wave along the body ( Figure 3B ) . There are relatively long periods of time where the shape changes by a continuous accumulation of the phase angle ( Figure 3C ) , and the speed of this rotation predicts the speed at which the worm crawls ( Figure 3D ) . In contrast to the first two modes , the third mode u3 ( s ) contributes to a nearly constant curvature throughout the middle half of the body ( Figure 2C ) . The distribution of the mode amplitude a3 has a long tail ( Figure 4A ) , and body shapes chosen from these tails ( Figure 4B ) exhibit the Ω configuration classically identified with large-angle turning behavior [10] . Large amplitudes of a3 also correspond to gradual turns in the worm trajectory along the agar ( Figure 4C ) . The fourth mode u4 ( s ) contributes to the shape of the head and tail region of the worm . Figure 2D shows that u4 ( s ) captures a large amount of the shape variance in those regions . Head movements of the worm are likely important in foraging [17] and navigation [18] . The emergence of a separate mode is likely due to the fact that head of C . elegans can move independently of the body and is controlled by a separate set of neck muscles . The connections between mode amplitudes and the motion of the worm along the agar—as in Figures 3D and 4C—are genuine tests of the functional meaning of our low dimensional description . Quite explicitly , our analysis of worm shapes is independent of the extrinsic coordinates and hence our definition of modes and amplitudes is blind to the actual position and orientation of the worm . Of course , in order to move the worm must change shape , and our description of the shape in terms of mode amplitudes captures this connection to movement . Thus , to crawl smoothly forward or backward the worm changes its shape by rotating clockwise or counterclockwise in the plane formed by the mode amplitudes a1 and a2; the speed of crawling is set by the speed of the rotation . Similarly , to change direction the worm changes shape toward larger magnitudes of the mode amplitude a3 , and we see this connection even without defining discrete turning events . The eigenworms provide a coordinate system for the postures adopted by C . elegans as it moves; to describe the dynamics of movement we need to find equations of motion in this low dimensional space . We start by focusing on the plane formed by the first two mode amplitudes a1 and a2 . Figure 3 suggests that within this plane the system stays at nearly constant values of the radius , so that the relevant dynamics involves just the phase angle φ ( t ) . To account for unobserved and random influences these equations need to be stochastic , and to support both forward and backward motion they need to form a system of at least second order . Such a system of equations would be analogous to the description of Brownian motion using the Langevin equation [19] , [20] . Thus we search for equations of the form ( 4 ) Here F[φ ( t ) , ω ( t ) ] defines the average acceleration as a function of the phase and phase velocity , by analogy to the force on a Brownian particle . The noise is characterized by a random function η ( t ) which we hope will have a short correlation time , and we allow the strength of the noise F[φ ( t ) , ω ( t ) ] to depend on the state of the system , by analogy to a temperature that depends on the position of the Brownian particle . In Figure 5A we show our best estimate of the mean acceleration F[φ , ω] ( see Materials and Methods for details ) . Once we know F , we subtract this mean acceleration from the instantaneous acceleration to recover trajectories of the noise , and the correlation function of this noise is shown in Figure 5B . The correlation time of the noise is short , which means that we have successfully separated the dynamics into two parts: a deterministic part , described by the function F[φ , ω] , which captures the average motion in the {a1 , a2} plane and hence the relatively long periods of constant oscillation , and a rapidly fluctuating part η ( t ) that describes “jittering” around this simple oscillation as well as the random forces that lead to jumps from one type of motion to another . We can imagine a hypothetical worm which has the same deterministic dynamics as we have found for real worms , but no noise . We can start such a noiseless worm at any combination of phase and phase velocity , and follow the dynamics predicted by Equation 4 , but with σ = 0 . These dynamics are diverse on short time scales , depending in detail on the initial conditions , but eventually all initial conditions lead to one of a small number of possibilities ( Figure 5C ) : either the phase velocity is always positive , always negative , or decays to zero as the system pauses at one of two stationary phases . Thus , underneath the continuous , stochastic dynamics we find four discrete attractors which correspond to well defined classes of behavior . We can compare the predicted behavioral states with the motion of real worms that include transitions between these states . Figure 5D is the joint probability density , ρ ( ω , φ ) , of worms sampled at 32 Hz; the trajectory of a single worm visiting all three predicted behavioral states is indicated by the overlay . The forward ( ω>0 ) and backward ( ω<0 ) motions match well with previously calculated attractor states , and pauses in the trajectory of real worms correspond to the calculated pause basins ( ω = 0 ) . Surprisingly , the transition between forward and backward motion is not arbitrary , but occurs most often along specific phase dependent trajectories . The behavior of C . elegans , particularly in response to sensory stimuli , traditionally has been characterized in probabilistic terms: worms respond by changing the probability of turning or reversing [17] , [21] , [22] . This randomness could reflect an active strategy on the part of the organism , or it could reflect the inability of the nervous system to distinguish reliably between genuine sensory inputs and the inevitable background of noise . Our description of motor behavior measured with high time resolution offers us the opportunity to revisit the “psychophysics” of C . elegans . We consider the response to brief ( 75 ms ) , small ( ΔT≈0 . 1°C ) changes in temperature , induced by pulses from an infrared laser ( see Materials and Methods ) . These stimuli are large enough to elicit responses [12] but well below the threshold for pain avoidance [16] . In Figure 6 we show the distribution ρt ( ω ) of phase velocities as a function of time relative to the thermal pulse . All of the worms were crawling forward at the moment of stimulation , so the initial phase velocities are distributed over a wide range of positive values . Within one second , the distribution narrows dramatically , concentrating near zero phase velocity . This behavior is consistent with the worm visiting the pause states described above in the deterministic dynamics , and may be similar to the pausing response seen when worms are subjected to mechanical stimuli [23] . Arrival in the pause state is stereotyped both across trials and across worms . By analogy with conventional psychophysical methods [1] , we can ask how reliably an observer could infer the presence of the heat pulse using the worm's response . We find that just measuring the phase velocity ω at single moment in time after the pulse is sufficient to provide ≈75% correct detection of this small temperature change in single trials . Our discussion thus far has separated the dynamics of the worm into two very different components: the {a1 , a2} plane with its phase dynamics , responsible for crawling motions , and the mode a3 , which is connected with large curvature turns . Because these modes are eigenvectors of a covariance matrix their instantaneous amplitudes are not linearly correlated , but this does not mean that the dynamics of the different motions are completely uncoupled . We found the clearest indications of mode coupling between the phase in the {a1 , a2} plane and the amplitude a3 at later times , which is illustrated by the correlation function in Figure 6B . The diagonal band of positive correlation reflects the phase dependent bending motions of normal crawling . This pattern of correlations is perturbed strongly by thermal stimuli ( t , t′>0 ) . The fact that the correlations between phase and the turning mode are stimulus dependent implies that the response of the turning mode to thermal stimuli depends on the phase which the worm finds itself at the time of the stimulus . Perhaps some of the apparent randomness of turning responses thus is related to the fact that repeated thermal stimuli catch the worm at different initial phases . To test this idea , Figure 6C shows the average response of a3 when worms are thermally stimulated with their head turned to either the dorsal or ventral side . Worms stimulated when making a ventral head swing ( −2≤φ≤−1 ) make bends in the dorsal direction ( a3<0 ) , and vice versa . Note that the thermal pulse itself does not have a handedness , so that if the pulses are not synchronized to the state of the worm there should be no systematic preference for dorsal vs . ventral handed turns . As a further test of this idea , we implemented our analysis online , allowing an estimate of the phase with a delay of less than 125 ms . We then deliver an infrared pulse when the phase falls within a phase window that corresponds to either dorsal– or ventral– directed head swings . The predicted consequence is that the worm should turn in the opposite direction to the laser stimulation , and is confirmed in Figure 6D .
Our central result is a new , quantitative , and low-dimensional description of C . elegans motor behavior . Conceptually similar results have been obtained for aspects of motor control in humans and other primates , where postures or trajectories of limbs , hands or eyes are confined to spaces of low dimensionality despite the potential for more complex motions [24]–[27] . For C . elegans itself , recent quantitative work has focused on simplifying behavior by matching to a discrete set of template behaviors , such as forward and backward motion of the center of mass [5] , sinusoidal undulations of the body [6] , or Ω bends [7] . Our results combine and generalize these ideas . Motor behaviors are described by projection of the body shape onto a small set of templates ( the eigenworms ) , but the strengths of these projections vary continuously . The templates are sinuous , but not sinusoidal , because the fluctuations in posture are not homogeneous along the length of the worm . Our description of shape is intrinsic to the worm and invariant to the center of mass position , but motion in shape space predicts the center of mass motion . There are discrete behavioral states , but these emerge as attractors of the underlying dynamics . Most importantly , our choice of four eigenworms is driven not by hypotheses about the relevant components of behavior , but by the data itself . The construction of the eigenworms guarantees that the instantaneous amplitudes along the different dimensions of shape space are not correlated linearly , but the dynamics of the different amplitudes are nonlinear and coupled; what we think of as a single motor action always involves coordinating multiple degrees of freedom . Thus , forward and backward motion correspond to positive and negative phase velocity in Figure 3 , but transitions between these behavioral states occur preferentially at particular phases . Similarly , turns involve large amplitude excursions along a3 , but motion along this mode is correlated with phase in the ( {a1 , a2} ) plane , and this correlation itself has structure in time ( Figure 6B ) . The problems of C . elegans motor control are simpler than for higher animals , but these nonlinear , coupled dynamics give a glimpse of the more general case . Perhaps because of the strong coupling between the turning mode a3 and the wriggling modes a1 , a2 , we have not found an equation of motion for a3 alone which would be analogous to Equation 4 for the phase . Further work is required to construct a fully three dimensional dynamics which could predict the more complex correlations such as those in Figure 6B . Turning should emerge from these equations not as another attractor , but as an ‘excitable’ orbit analogous to the action potential in the Hodgkin–Huxley equations or to recent ideas about transient differentiation in genetic circuits [28] . A major challenge would be to show that the stochastic dynamics of these equations can generate longer sequences of stereotyped events , such as pirouettes [29] . We have shown that a meaningful set of behavioral coordinates can uncover deterministic responses . A response might seem stochastic or noisy because it depends on one or more behavioral variables that are not being considered . In our experiments , nonlinear correlations among the behavioral variables suggest that some of the randomness in behavioral responses could be removed if sensory stimuli are delivered only when the worm is at a well defined initial state , and we confirmed this prediction by showing that phase–aligned thermal stimuli can ‘steer’ the worm into trajectories with a definite chirality . A crucial aspect of these experiments is that the stimulus is scalar—a temperature change in time has no spatial direction or handedness—but the response , by virtue of the correlation between stimulus and body shape , does have a definite spatial structure . The alignment of thermal stimuli with the phase of the worm's movement in these experiments mimics the correlation between body shape and sensory input that occurs as the worm crawls in a thermal gradient , so the enhanced determinism of responses under these conditions may be connected to the computations which generate nearly deterministic isothermal tracking [22] , [30] . More generally , all behavioral responses have some mixture of deterministic and stochastic components . In humans and other primates , it seems straightforward to create conditions that result in highly reproducible , stereotyped behaviors , such as reaching movements [31] . At the opposite extreme , bacterial motility is modulated in response to sensory inputs , but these responses seem fundamentally probabilistic [32] . Some of these differences may result from the physical nature of sensory stimuli in organisms of vastly different size [33] , [34] , but some of the differences may also result from differences of strategy or available computational power . The more stochastic the response , the more challenging it is to characterize behavior quantitatively and to link behavior with underlying molecular and neural components , as is clear from recent work on Drosophila olfaction ( see , for example , [35] ) . We hope that our approach to the analysis of behavior may help to uncover more deterministic components of the sensory–motor responses in other model organisms . More than forty years of work on C . elegans has led to a fully sequenced genome [36] and to the complete wiring diagram of the nervous system [37] . Significant steps have been made toward the original dream [38] of connecting genes , neurons , and behavior [11] , [39] , [40] . Nonetheless , with the advances in molecular , cellular , and circuit analyses , our ability to probe the mechanisms which generate behavior substantially exceeds our ability to characterize the behavior itself . Perhaps our work provides a step toward addressing this imbalance .
The imaging system consists of a Basler firewire CMOS camera ( A601f , Basler , Ahrensburg , Germany ) with 4x lens ( 55–901 , Edmund Optics , Barrington , NJ ) and a fiber optic trans-illuminator ( DC-950 , Dolan-Jenner , Boxborough , MA ) mounted to an optical rail ( Thorlabs , Newton , NJ ) . The rail is attached to a XY translation stage ( Deltron , Bethel , CT ) which is driven by stepper motors ( US Digital , Vancouver , Washington ) . The stage driver is a homemade unit utilizing a SimpleStep board ( SimpleStep , Newton , NJ ) and Gecko stepper motor drivers ( Geckodrive , Santa Ana , CA ) . Image acquisition , processing , and stage driver control was done using LabVIEW ( National Instruments , Austin , TX ) . Images of worms were isolated and identified using the image particle filter . A raw unprocessed JPEG image and a filtered process binary PNG image were written to the hard drive at rates up to 32 Hz . Concurrently at 4 Hz , the center of mass of the worm was calculated and the distance from the center of the field of view in pixels was computed . An error signal was then calculated via a coordinate transformation between the camera reference frame and the translational stage reference frame and the XY stage was moved to center the worm in the field of view . The C . elegans strain , N2 , was grown at 20°C and maintained under standard conditions [41] . Before each experiment , excess moisture from NGM assay plates ( 1 . 7% Bacto Agar , 0 . 25% Bacto-Peptone , 0 . 3% NaCl , 1 mM CaCl2 , 1 mM MgSO4 , 25 mM potassium phosphate buffer , 5 µg/mL cholesterol ) was removed by leaving them partially uncovered for 1 hr . A copper ring ( 5 . 1-cm inner diameter ) pressed into the agar surface prevented worms from crawling to the side of the plate . Young adults were rinsed of E . coli by transferring them with a worm pick from OP50 bacterial food plates into NGM buffer ( same inorganic ion concentration as NGM assay plates ) and letting them swim for 1 minute . Worms were transferred from the NGM buffer to the center of the assay plate ( 9-cm Petri dish ) . The location of the dorsal side of the worm was noted via a stereomicroscope . The plates were covered and tracking began after 1 minute and lasted no longer than 60 minutes . In the rare cases where worms stopped moving before the completion of the run , the data were excluded . Images of worms captured by the worm tracker were processed using MATLAB ( Mathworks , Natick , MA ) . Cases of self-intersection were excluded from processing . Images of worms were thinned to a single-pixel-thick backbone , and aligned so that the dorsal/ventral directions were consistent . A spline was fit through these points and then discretized into 101 segments , evenly spaced in units of the backbone arclength . The N = 100 angles between these segments were calculated and an overall rotation mode was removed by subtracting ∑θ ( s ( i ) ) /N from each angle . The shape covariance matrix C ( s , s′ ) = 〈 ( θ ( s ) –〈θ〉 ) ( θ ( s′ ) –〈θ〉 ) 〉 was constructed from 9 freely crawling worms sampled at 4 Hz , for a period of 30 minutes ( a total of 60 , 000 images ) . Each eigenworm uμ ( s ) is an eigenvector of the covariance matrix ∑s′ C ( s , s′ ) uμ ( s′ ) = λμuμ ( s ) . The fractional variance captured by K eigenvectors is thus , where σ2 = Σμλμ is the total variance of the measurements . The same eigenworms shown in Figure 2 were used throughout the various analysis reported in the paper . The worm's phase was defined as φ = tan−1 ( −a2/a1 ) where a1 and a2 were both normalized to unit variance . The crawling speed was defined as the time derivative of the worm's center of mass . For the analysis of phase dynamics we sampled the worm shape at 32 Hz . Data for the construction of the equations of motion came from 12 worms , 5 trials per worm , with 4000 frames per trial . We also filtered each mode time series through a low-pass polynomial filter so that for each frame ( 26≤m≤3974 ) , where {pj} are the best-fit polynomial coefficients . Mode time derivatives were calculated using derivatives of the polynomial filter . None of our results depend critically on the properties of the filter . The Langevin equations governing the phase dynamics are shown Eq . ( 4 ) and we learn the functions {F ( φ , ω ) , σ ( φ , ω ) } directly from the time series [42] , [43] . By construction 〈σ[φ ( t ) , ω ( t ) ]η ( t ) 〉 = 0 and therefore the optimal rms estimate of F ( φ , ω ) is the conditional mean . We estimate F by assuming a functional expansion , where the model parameters {} were determined by minimizing the rms error on training data ( 90% ) and the hyperparameters {mmax = 5 , pmax = 5} were chosen to minimize error on held-out data ( 10% ) . Once F is known we can determine the noise in the system; we normalize 〈η2〉 so that . The attractors contained within our derived dynamics were obtained by evolving initial conditions spanning the sampled {ω , φ} plane for long times ( 93 . 75 s≈47 cycles ) . In the deterministic dynamics all trajectories evolve to one of four asymptotic states and we observed no switching . Worms were prepared as described earlier but raised at a lower temperature ( 17°C ) leading to a lower average ω before the thermal stimulus . A collimated beam with a 1/e diameter of 5 . 6 mm ( standard stimulus ) or 1 . 5 mm ( painful ) from a 1440 nm diode laser ( FOL1404QQM , Fitel , Peachtree City , GA ) was positioned to heat the area covering the worm . The diode laser was driven with a commercial power supply and controller ( Thorlabs , Newton , NJ ) . Power and duration of the beam was controlled through software using LabVIEW . For each worm , 1000 seconds of data was collected in cycles of 50 seconds . 12 . 5 seconds into each cycle the laser was turned on for a duration of 75 ms at 150 mW ( standard ) or 250 ms at 100 mW ( painful ) . The temperature increase caused by the laser pulses was measured using a 0 . 075mm T-type thermocouple ( coco-003 , Omega , Stamford , CT ) placed on the surface of the agar and sampled with a thermocouple data acquisition device ( USB-9211 , National Instruments ) . For each measurement , 60 trials of 30 s cycles were averaged . The temperature increase was calculated by subtracting the maximum temperature ( recorded immediately after the laser pulse ) from the baseline temperature ( recorded 9 s after the laser pulse ) . The temperature increase for the standard pulse was 0 . 12°C and the increase of the painful pulse was 0 . 73°C . Data were taken from a collection of 13 worms , each stimulated with 20 repetitions of a ΔT = 0 . 1°C pulse . In Figure 6A , the time-dependent probability density ρt ( ω ) was smoothed before the onset of the impulse with a gaussian low-pass filter of size 0 . 19 s in the t direction and 0 . 17 cycles/s in the ω direction . In Figure 6B the correlation function was calculated as follows . Far from the time of the impulse ( frames 800 to 1574 , impulse on frame 400 ) , we expect time-translation invariance Cpost ( t , t′ ) = g ( t–t′ ) +ξpost ( t , t′ ) where g ( Δ ) = 〈C ( i , j ) 〉i–j = Δ is the true correlation function and ξpost ( t , t′ ) characterizes statistical error . Similarly in a time window around the impulse ( frames 24 to 800 ) , Cstim ( t , t′ ) = g ( t–t′ ) +ξstim ( t , t′ ) . However , the thermal impulse breaks this invariance and ξstim ( t , t′ ) contains both sampling fluctuations and stimulus-dependent correlation dynamics . To separate these effects we use singular value decomposition to compare ξpost and ξstim . We write each matrix ξpost/stim ( t , t′ ) = ∑t″Upost/stim ( t , t″ ) Spost/stim ( t″ , t″ ) Vpost/stim ( t″ , t′ ) and find that only two singular values of ξstim are significantly larger than ξpost . We then reconstruct the two-point function around the stimulus as . Preparation of worms and instrumentation were the same as described for the thermal impulse response . However , instead of processing worm images off-line , real-time calculation of the eigenworms and shape phase φ was done using custom dynamic-linked image processing libraries written in C along with supporting LabVIEW code . The modes were computed as previously described except that the spline interpolation algorithm was replaced with a Hermitian interpolation algorithm to reduce the processing time . The processing time was short enough to simultaneously track and calculate modes at 8 Hz . For phase dependent measurements , the laser was fired when the worm was moving forward and φ fell within a prescribed interval ( width 1 radian ) . The laser pulse ( 150 mW ) lasted for 75 ms and caused a temperature increase of 0 . 12°C . For each run a pair of triggering phase windows ( 0 to −1 , and 2 . 1 to 3 . 1 radians ) corresponding to the dorsal- and ventral-directed head swing was used . The sequence of each run started with a 5 minute period of no stimulus followed by the pair of phase dependent stimuli . The order of each pair of stimulus conditions was switched for each successive run . For the randomized pulse control experiments , the laser was fired with a uniform phase probability , but with conditions that restricted the firing interval to be longer than 2 seconds . The time-average change in orientation of the worm's path , 〈〉 ( rad/s ) , was calculated from the angular changes between the positions of the center of mass of the worm during forward runs of at least 4 s in length . Given positions ( r1 , r2 , r3 , … , rN ) , the angles between connecting segments ( r2– r1 , r3– r2 , r4– r3 , … , r ( N-1 ) ) were calculated . 〈〉 was calculated in intervals of 10 s . Since the distributions were Gaussian ( data not shown ) with similar variance , we used the Student's t-test to determine if the values of under thermal stimulation were significantly different than the control ( p<0 . 05 ) . Since we were interested in the change in orientation during forward motion we excluded trajectory data that contained large turns or reversals along with angular changes greater than π/4 radians . These events were automatically detected by measuring the compactness of the worm shape . Compactness was calculated by measuring the longest distance between two points in the worm shape ( also known as the max feret distance ) and normalizing this with the maximum value for the entire data run . Turns were flagged when the compactness fell below 0 . 6 . | A great deal of work has been done in characterizing the genes , proteins , neurons , and circuits that are involved in the biology of behavior , but the techniques used to quantify behavior have lagged behind the advancements made in these areas . Here , we address this imbalance in a domain rich enough to allow complex , natural behavior yet simple enough so that movements can be explored exhaustively: the motions of Caenorhabditis elegans freely crawling on an agar plate . From measurements of the worm's curvature , we show that the space of natural worm postures is low dimensional and can be almost completely described by their projections along four principal “eigenworms . ” The dynamics along these eigenworms offer both a quantitative characterization of classical worm movement such as forward crawling , reversals , and Omega-turns , and evidence of more subtle behaviors such as pause states at particular postures . We can partially construct equations of motion for this shape space , and within these dynamics we find a set of attractors that can be used as a rigorous definition of behavioral state . Our observations of C . elegans reveal a precise and complete language of motion and new aspects of worm behavior . We believe this is a lesson with promise for other organisms . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"neuroscience/behavioral",
"neuroscience",
"biophysics",
"neuroscience/motor",
"systems"
]
| 2008 | Dimensionality and Dynamics in the Behavior of C. elegans |
A recombinant severe acute respiratory syndrome coronavirus ( SARS-CoV ) lacking the envelope ( E ) protein is attenuated in vivo . Here we report that E protein PDZ-binding motif ( PBM ) , a domain involved in protein-protein interactions , is a major determinant of virulence . Elimination of SARS-CoV E protein PBM by using reverse genetics caused a reduction in the deleterious exacerbation of the immune response triggered during infection with the parental virus and virus attenuation . Cellular protein syntenin was identified to bind the E protein PBM during SARS-CoV infection by using three complementary strategies , yeast two-hybrid , reciprocal coimmunoprecipitation and confocal microscopy assays . Syntenin redistributed from the nucleus to the cell cytoplasm during infection with viruses containing the E protein PBM , activating p38 MAPK and leading to the overexpression of inflammatory cytokines . Silencing of syntenin using siRNAs led to a decrease in p38 MAPK activation in SARS-CoV infected cells , further reinforcing their functional relationship . Active p38 MAPK was reduced in lungs of mice infected with SARS-CoVs lacking E protein PBM as compared with the parental virus , leading to a decreased expression of inflammatory cytokines and to virus attenuation . Interestingly , administration of a p38 MAPK inhibitor led to an increase in mice survival after infection with SARS-CoV , confirming the relevance of this pathway in SARS-CoV virulence . Therefore , the E protein PBM is a virulence domain that activates immunopathology most likely by using syntenin as a mediator of p38 MAPK induced inflammation .
Severe acute respiratory syndrome coronavirus ( SARS-CoV ) was identified as the etiological agent of a respiratory disease that emerged in Southeast China at the end of 2002 . SARS-CoV spread to more than 30 countries within six months , infecting 8000 people with an average mortality of 10% [1] . After July 2003 , only a few community and laboratory-acquired cases have been reported ( http://www . who . int/csr/sars/en/ ) . Nevertheless , coronaviruses , including those similar to the strain that caused the epidemic , are widely disseminated in bats circulating all over the world , making a future outbreak possible [2]–[5] . In September 2012 , a novel coronavirus , named Middle East respiratory syndrome coronavirus ( MERS-CoV ) was identified in two persons with severe respiratory disease [6] , [7] . By now , 701 laboratory-confirmed MERS-CoV cases , including 249 deaths , have been diagnosed in several countries ( http://www . who . int/csr/don/2014_06_16_mers/en/ ) . Most patients reported respiratory disease symptoms , occasionally accompanied by acute renal failure [8] . A better understanding of the molecular mechanisms underlying the virulence of these highly pathogenic coronaviruses will facilitate the development of therapies to alleviate or prevent the impact of coronavirus infection on human health . SARS-CoV belongs to the Coronavirinae subfamily , genus β and is an enveloped virus with a single-stranded positive sense 29 . 7 kb RNA genome [9] . SARS-CoV envelope ( E ) protein is a small integral membrane protein of 76 amino acids that contains a short hydrophilic amino-terminus followed by a hydrophobic region , and a hydrophilic carboxy-terminus [10] . The hydrophobic region forms at least one amphipathic α-helix that oligomerizes to form an ion-conductive pore in membranes [11]–[13] . E protein is present within virions in very small amounts , however it is abundant in the infected cells [14] , and it is mainly localized in the endoplasmic reticulum Golgi intermediate compartment ( ERGIC ) , where it actively participates in virus budding , morphogenesis and trafficking [15]–[17] . Interestingly , SARS-CoV lacking the E protein was attenuated in different animal models , such as hamsters , transgenic mice that expressed the SARS-CoV receptor , human angiotensin converting enzyme 2 ( hACE-2 ) , and conventional mice using a mouse adapted SARS-CoV [18]–[21] , indicating that SARS-CoV E gene may be a virulence factor . We have previously shown that SARS-CoV E protein increased the apoptosis and reduced the stress response induced after SARS-CoV infection [22] . Transient expression of SARS-CoV E protein in trans showed that the protein associated with Caenorhabditis elegans lin-7 protein 1 ( PALS1 ) , a tight junction-associated protein , is an E protein interacting partner [23] . PALS1 bound E protein through the post-synaptic density protein-95/discs Large/zonula occludens-1 ( PDZ ) domain of PALS1 [23] , which recognized the last four carboxy-terminal amino acids of E protein that form a type II PDZ-binding motif ( PBM ) with the consensus sequence X-φ-X-φCOOH ( where X represents any amino acid and φ is a hydrophobic residue , usually V , I or L ) [24] . However , the relevance of this interaction during virus infection and its impact on virulence in vivo was not tested . PDZ domains are protein–protein recognition sequences , consisting of 80–90 amino acids that bind to a specific peptide sequence ( PBM ) , usually located at the end of the carboxy-terminus of a target protein [25]–[27] . Proteins containing PDZ domains are typically found in the cell cytoplasm or in association with the plasma membrane and play a role in a variety of cellular processes of significance to viruses , such as cell-cell junctions , cellular polarity , and signal transduction pathways [28] . PDZ domains are found in thousands of proteins and are widespread in eukaryotes and eubacteria [29] . Just in the human genome , there are more than 900 PDZ domains in at least 400 different proteins [30] . These protein-protein interactions modulate cellular pathways influencing viral replication , dissemination in the host or pathogenesis [28] . As previously described , SARS-CoV E protein contains a PBM [23] . However , the relevance of this motif in the context of infection and its role in virus pathogenesis has not been elucidated . In this study , we have identified the SARS-CoV E protein PBM as a virulence determinant in vivo . Infection with recombinant viruses lacking an E protein PBM were attenuated in mice , which was accompanied by a decreased expression of inflammatory cytokines during infection , and a substantial increase of survival . In contrast , all mice infected with viruses containing E protein PBM died . We further found that the E protein PBM interacted with the cellular protein syntenin during SARS-CoV infection , affecting p38 mitogen-activated protein kinase ( MAPK ) activation , a protein involved in the expression of inflammatory cytokines , responsible of the pathogenicity associated to SARS-CoV infection . In addition , mice treated with a p38 MAPK inhibitor showed a significant increase in survival after infection with SARS-CoV . Together , our findings provide novel insights into how highly pathogenic viruses , such as SARS-CoV , induce virulence , suggesting potential therapeutic targets to improve the prognosis in patients .
To evaluate the role of SARS-CoV E protein PBM in virus pathogenesis , a set of recombinant SARS-CoVs with E protein PBM mutated or truncated ( SARS-CoV-E-PBMs ) were generated using an infectious cDNA encoding a mouse adapted ( MA15 ) SARS-CoV [31] , [32] . Infection of BALB/c mice with SARS-CoV-MA15 causes morbidity , mortality and pulmonary pathology , similar to the symptoms observed in human SARS [31] . In SARS-CoV-E-ΔPBM , abbreviated as ΔPBM , the last 9 amino acids of E protein were deleted , truncating the carboxy-terminus , and eliminating the PBM ( Figure 1A ) . In contrast , the PBM was abolished in SARS-CoV-E-mutPBM ( mutPBM ) by mutating the last 4 amino acids to glycine , maintaining the full-length SARS-CoV E protein . In the last recombinant , termed SARS-CoV-E-potPBM ( potPBM ) , four amino acids within E protein were replaced by alanine , modifying the E protein carboxy-terminal sequence while maintain the consensus PBM residues ( Figure 1A ) . To test whether mutation or deletion of SARS-CoV E protein PBM alters virus fitness in vitro , growth kinetics of SARS-CoV-E-PBM mutants were analyzed in comparison to the parental virus ( wt ) and the virus lacking the full-length E protein SARS-CoV-ΔE ( ΔE ) in monkey Vero E6 and mouse DBT-mACE2 cells [33] ( Figure 1B ) . Despite the observed replication defects of the ΔPBM and mutPBM viruses at 24 hpi in DBT-mACE2 cells , the parental virus including native E protein or mutants lacking a PBM reached similar titers at 72 hpi , both in Vero E6 cells and in DBT-mACE2 cells ( Figure 1B ) . In contrast , the titer of ΔE virus was reduced around 50-fold ( Figure 1B ) . This result indicated that SARS-CoV E protein PBM was not essential for efficient virus growth in Vero E6 cells and , at late times post infection , in DBT-mACE2 cells . To analyze the pathogenicity of SARS-CoV-E-PBM mutants , 16 week-old female BALB/c mice were intranasally inoculated with recombinant viruses . Body weight ( Figure 2A ) and mortality ( Figure 2B ) of each mouse were monitored daily . Mock-infected mice and those infected with a virus lacking E protein , did not lose weight and all survived . In contrast , mice infected with recombinant viruses including an E protein PBM , either the original PBM ( wt ) or a potential PBM ( potPBM ) , underwent severe weight loss ( Figure 2A ) and developed signs of illness , including shaking , ruffling of the fur , and loss of mobility , resulting in 100% mortality by 9 days post infection ( dpi ) ( Figure 2B ) . Interestingly , mice infected with the viruses in which E protein PBM was abolished ( mutPBM ) or deleted ( ΔPBM ) , showed moderate or no weight losses , respectively , and 100% survival in both cases ( Figure 2B ) . The fact that ΔPBM is more attenuated that mutPBM suggests the presence of sequences , outside PBM , further contributing to virus pathogenesis . These results indicated that elimination of E protein PBM led to virus attenuation in vivo and that the presence of a functional PBM conferred virulence , as mutant potPBM in which 4 amino acids in the carboxy-terminal domain were mutated to alanine , conserving consensus residues in the PBM was still virulent . To evaluate the effect of the E protein PBM in virus growth in vivo , BALB/c mice were intranasally inoculated with recombinant viruses , and viral titers in the lung were determined at 2 and 4 dpi ( Figure 2C ) . Viruses ΔE and ΔPBM showed decreased titers in lungs at both 2 and 4 dpi , as compared with the wt or potPBM viruses . Mutant ΔE replicated at a higher level than ΔPBM at 4 dpi in lungs of infected mice , suggesting that the ΔPBM E protein may induce an antiviral state after infection . Interestingly , the virus lacking a functional E protein PBM but conserving a full-length E protein ( mutPBM ) grew to similar levels as the wt virus at 2 and 4 dpi , indicating that a virus lacking a canonical PBM in E protein , but conserving E protein full-length displayed an attenuated phenotype , although it efficiently replicated in vivo . To analyze the mechanisms by which E protein PBM confers virulence in vivo , lung pathology was examined in infected BALB/c mice at 2 and 4 dpi . At these time points , no obvious gross lesions or changes in weight were observed in the lungs of non-infected mice or in those from mice infected with viruses lacking functional E protein PBM ( ΔE , ΔPBM and mutPBM ) . In contrast , at 2 and especially at 4 dpi , lungs of mice infected with SARS-CoVs with an E protein containing a functional PBM ( wt and potPBM ) were highly edematous , with profuse hemorrhagic areas , leading to significant lung weight increase at 4 dpi , probably due to leukocyte infiltration ( Figures 3A and 3B ) . To further characterize the pathology induced in BALB/c mice by the infection with SARS-CoVs with and without E protein PBM , lung sections were collected at 2 and 4 dpi , stained with hematoxylin and eosin ( Figure 3C ) and pulmonary histopathology scores for edema and cellular infiltrates were quantified according to previously described procedures [34] ( Figures 3D and 3E ) . Histological examination of lungs from mock or SARS-CoV-ΔE-infected mice showed minimal evidence of damage or cellular infiltration at 2 and 4 dpi ( Figure 3C , 3D and 3E ) . In contrast , mice infected with recombinant viruses containing functional E protein PBM ( wt and potPBM ) revealed interstitial and peribronchial cell infiltration and edema in both alveolar and bronchiolar airways at 2 and , mainly , at 4 dpi ( Figure 3C , 3D and 3E ) . Interestingly , mice infected with viruses containing E protein but lacking functional PBM sequences ( ΔPBM and mutPBM ) , showed minimal epithelial damage or lung edema and only small amounts of inflammatory cell infiltrates at 4 dpi ( Figure 3C , 3D and 3E ) . These data indicated that the attenuation observed for viruses lacking E protein PBM correlated with decreased lung pathology . The effect of the presence of a PBM in SARS-CoV E protein on host gene expression during BALB/c mice infection was analyzed using microarrays at 2 dpi . MIAME-compliant results of the microarrays have been deposited in the Gene Expression Omnibus database ( GEO [National Center for Biotechnology Information] , accession code GSE52920 ) . A total of 922 and 640 cellular genes were differentially expressed in lung of mice infected with SARS-CoV with ( wt ) or without functional E protein PBM ( mutPBM ) as compared with mock-infected mice ( Figure 4A ) . Remarkably , 319 genes were differentially expressed in mice infected with mutPBM compared to wt , despite a difference at only 4 amino acid positions between the two viruses . Of these , 218 genes were found to be upregulated and 101 genes were downregulated ( Figure 4A ) . Interestingly , analysis using DAVID software [35] revealed that most of the downregulated genes in mutPBM versus wt infections clustered within wound response and inflammatory and defense response pathways ( Figure 4B ) . The most significant genes present in at least one of these groups were serum amyloid A2 ( SAA2 ) , chemokine ( C-C motif ) ligand 3 ( CCL3 ) , chemokine ( C-X-C motif ) ligand 1 ( CXCL1 ) , chemokine ( C-X-C motif ) ligand 5 ( CXCL5 ) , calcitonin ( CALCA ) , serum amyloid A1 ( SAA1 ) , chemokine ( C-X-C motif ) ligand 10 ( CXCL10 ) , chemokine ( C-C motif ) ligand 2 ( CCL2 ) , interleukin 1 beta ( IL1B ) , orosomucoid 1 ( ORM1 ) , interleukin 6 ( IL6 ) , chemokine ( C-C motif ) ligand 4 ( CCL4 ) and chemokine ( C-X-C motif ) ligand 9 ( CXCL9 ) ( Figure 4C ) . The differential expression of a group of cellular genes identified in the microarray ( CXCL10 , CCL2 and IL6 ) was confirmed by RT-qPCR analysis using RNA from the lungs of mice infected with all the recombinant viruses generated , collected at 2 dpi , in relation to RNAs from mock-infected mice . 18S ribosomal RNA ( rRNA ) was used to normalize the data [36] , [37] ( Figure 4D ) . Using both microarray and RT-qPCR , we identified genes differentially expressed in the lungs of mice infected with viruses with or without E protein PBM ( Figures 4C and 4D ) . Viruses lacking E protein PBM induced a decreased expression of inflammatory cytokines . These data indicated that the exacerbated host innate immune response triggered during SARS-CoV infection was reduced in the absence of SARS-CoV E protein PBM , which may explain the attenuated phenotype of these viruses . SARS-CoVs defective in E protein PBM presented an attenuated phenotype that correlated with a decreased inflammatory response . The absence of this motif may likely imply changes in interaction patterns with cellular proteins that may explain their reduced virulence . To identify these cellular factors , a yeast two-hybrid screen was performed using the carboxy-terminal domain of SARS-CoV E protein , where amino acids 36–76 of E protein carboxy-terminus ( ECT ) were used as bait ( Figure 5A ) . A random-primed cDNA library from human lung was screened . One of the most prominent results of the screening was the interaction between ECT and the syndecan binding protein ( syntenin ) ( Figure 5B ) , with a total of 13 positive clones corresponding to this protein ( GenBank accession number NM_005625 . 3 ) . The interaction was classified with a high confidence score ( predicted biological score of B ) [38] . Syntenin is a 32 kDa protein composed of a 113 amino acid N-terminal domain ( NTD ) with no obvious structural motifs , followed by two adjacent tandem PDZ domains ( PDZ1 and PDZ2 ) , that could mediate its interaction with E protein and a short 24 amino acid C-terminal domain ( CTD ) [39] ( Figure 5B ) . Interestingly , all 13 recovered syntenin cDNAs interacting with the E protein carboxy-terminus identified in the yeast two-hybrid platform contained the same two PDZ domains present in the cellular syntenin . To determine whether the cellular protein syntenin is also associated with SARS-CoV E protein in infected cells , and whether this interaction was mediated through the E protein PBM , Vero E6 cells were transfected with a plasmid encoding an N-terminal HA-tagged syntenin and then infected with recombinant viruses with or without E protein PBM . Non-transfected cells infected with wt virus were used as control . Syntenin or E protein were immunoprecipitated using extracts of infected cells using an HA specific monoclonal antibody or an E protein polyclonal antibody , respectively . Inmunoblot analysis using E and HA specific antibodies revealed that E protein coprecipitated with syntenin in cells infected with recombinant viruses containing the wt or potential E protein PBM ( Figure 5C ) . In contrast , E protein did not coprecipitate with syntenin in cells infected with viruses lacking the E protein or its PBM ( Figure 5C ) . These results indicate that the last 4 amino acids of SARS-CoV E protein form a functional PBM that , in the context of virus infection , mediate its association with the cellular protein syntenin . The absence of this interaction could be playing a role in the attenuation observed in viruses lacking E protein PBM . To evaluate whether SARS-CoV E protein and syntenin colocalize during infection and to determine if syntenin localization was altered during SARS-CoV infection , mock-infected Vero E6 cells and cells infected with the wt or mutPBM virus were analyzed by confocal immunomicroscopy using specific antibodies against the cellular protein syntenin and the SARS-CoV nucleocapsid ( N ) and E proteins ( Figure 6A ) . Syntenin was predominantly present in the nucleus of mock-infected cells . Upon wt infection , E protein was mainly localized at perinuclear regions as previously described [15] . Interestingly , after infection , syntenin partially colocalized with E protein in the perinuclear region and also relocalized to locations close to the plasma membrane ( Figure 6A ) . Furthermore , infection with the mutPBM virus led to N protein cytoplasmic localization as previously described [40] , and to a decrease in the relocalization of syntenin to the cytoplasm as compared with the parental virus . To determine whether E protein was involved in syntenin relocalization during SARS-CoV infection , Vero E6 cells were transiently transfected with an empty plasmid or a plasmid expressing E protein and both , syntenin and E protein , were detected with specific antibodies ( Figure 6B ) . As previously described with mock-infected cells , syntenin was mainly detected in the nucleus of cells transfected with an empty plasmid . However , in cells transfected with the plasmid expressing E protein , syntenin colocalized with E protein in the perinuclear region and adopted a distribution close to the plasma membrane ( Figure 6B ) . Furthermore , the percentage of mock-infected versus virus-infected cells that showed cytoplasmic accumulation of syntenin was quantified ( Figure 6C ) . Syntenin accumulated in the cytoplasm of 98 . 5% of the cells infected with the parental virus , whereas 31 . 2% of the mock-infected cells displayed syntenin in the cytoplasm . In cells infected with the mutPBM virus only 51 . 5% showed syntenin in the cytoplasm . These results indicated that syntenin partially colocalized with SARS-CoV E protein , and that it was redistributed from the nucleus to perinuclear regions , where E protein is accumulated , and also to regions close to the plasma membrane . Syntenin has been described as an important scaffolding protein that can initiate a signaling cascade resulting in the induction of p38 MAPK [41] . In this model , after its interaction with the extracellular matrix , syntenin induces phosphorylation and therefore , activation of p38 MAPK , a protein involved in the expression of proinflammatory cytokines [42] , [43] . To determine whether p38 MAPK was differentially activated in the lungs of mice infected with recombinant SARS-CoV with or without E protein PBM , 16 week-old female BALB/c mice were intranasally inoculated with these viruses . The activation of p38 MAPK was studied by Western blot analysis at 2 dpi , using a phospho-p38 MAPK ( p-p38 ) specific antibody to detect the active form , and an antibody specifically recognizing the total endogenous p38 MAPK . Actin served as loading control . Interestingly , the levels of active p38 MAPK were increased in the lungs of mice infected with SARS-CoV containing E protein PBM , compared to those found in lungs of mice infected with viruses lacking E protein PBM ( Figures 7A and 7C ) . To reinforce the data , p38 MAPK activation was studied in infected cells . To this end , Vero E6 cells were mock-infected or infected with recombinant viruses with an E protein with or without a PBM . Then , p38 MAPK activation was analyzed by Western blot at 24 hpi . Interestingly , an increase in p38 MAPK activation was observed during infection with viruses containing E protein PBM , similarly to what was observed in the lungs of SARS-CoV-infected mice ( Figures 7B and 7D ) . These results indicated that the E protein PBM is involved in p38 MAPK activation in response to SARS-CoV infection . We have shown above that SARS-CoV E protein PBM interacted with syntenin , and that infection with SARS-CoVs containing an E protein with a functional PBM led to an increase in p38 MAPK activation . As syntenin promotes p38 MAPK activation [41] , we hypothesized that syntenin relocalization from nucleus to cytoplasm during infection with SARS-CoV , containing an E protein including the PBM , may be responsible for the activation of the p38 MAPK pathway . To test this hypothesis , Vero E6 cells were mock-infected or infected with recombinant SARS-CoVs including an E protein with ( wt ) or without ( mutPBM ) E protein PBM . At 24 hpi , the cytosolic and nuclear fractions from SARS-CoV infected cells were collected , and the levels of syntenin and extent p38 MAPK activation in both fractions were determined by Western blot analysis using specific antibodies for syntenin and the non-phosphorylated and phosphorylated forms of p38 MAPK . The levels of histone H3 , total p38 MAPK and actin were used as controls . Syntenin levels in the cytosol fraction were increased during wt infection . Interestingly , mutPBM virus retained the ability to mislocalize a substantial amount of syntenin to the cytoplasmic fraction , possibly due to the ability of E protein to bind SARS-CoV 3a protein , which also contains a PBM , in addition to the one present in the E protein [44] , [45] . Furthermore , the presence of syntenin in the cytosol correlated with the activation of p38 MAPK ( Figure 8A ) . To determine whether syntenin relocalization from nucleus to cytoplasm mediated p38 MAPK activation , Vero E6 cells were transfected with an empty plasmid or a plasmid expressing human syntenin , and presence of this protein in the nucleus or the cytoplasm of the infected cells and the levels of p38 MAPK phosphorylation were studied by Western blot analysis using specific antibodies . The levels of histone H3 , total p38 MAPK and actin were used as controls . The results showed the presence of syntenin in the nucleus of cells transfected with both plasmids . In contrast , syntenin was only detected in the cytoplasmic fraction when the syntenin was overexpressed , and failed to accumulate in the nuclear fraction . This observation could be explained by the previously reported saturation of the nuclear import machinery , which leads to cytoplasmic retention of overexpressed proteins , as described for other proteins [46] . Interestingly , the presence of syntenin in the cytoplasm correlated with an increase of p38 MAPK activation ( Figure 8B ) . To further confirm the role of syntenin in the activation of p38 MAPK during infection by SARS-CoV with an E protein containing a functional PBM , siRNAs specifically designed to inhibit syntenin expression were used in mock-infected cells or in cells infected with the parental virus . p38 MAPK activation was analyzed by Western blot . Vero E6 cells were transfected twice by reverse transfection with either 25 nM of a validated negative control siRNA ( NEG ) or with similar amounts of each of two different siRNAs targeting endogenous syntenin . At 24 hpt , cells were mock-infected or infected with the parental virus at an MOI of 0 . 3 . At 24 hpi , syntenin mRNA levels were significantly ( 60 to 70% ) reduced in syntenin-silenced cells in relation to the cells transfected with a validated negative-control siRNA , as determined by qRT-PCR using specific Taqman gene expression assays ( Figure 8C ) . Accordingly , syntenin levels evaluated by Western blot analysis were also significantly reduced in syntenin-silenced cells . Moreover , this silencing was found to have a higher apparent impact on the reduction of this protein in the cytoplasm , probably because that this protein accumulates more efficiently in the nucleus than the cytoplasm after protein expression . Interestingly , inhibition of syntenin expression was accompanied by a decreased in p38 MAPK activation during the infection with the parental virus ( Figure 8D ) , whereas no changes in virus titers were observed ( Figure 8E ) . Overall , these results support the hypothesis that the interaction of E protein PBM with syntenin facilitates the recruitment of syntenin in the cytosol and leads to p38 MAPK activation . To analyze the contribution of SARS-CoV E protein PBM-mediated p38 MAPK activation to the disease observed during SARS-CoV infection in mice , 16 week-old female BALB/c mice were intraperitoneally administered with a control buffer or SB203580 , a highly specific inhibitor of p38 MAPK , and were mock-infected or infected with the rSARS-CoV-MA15 virus . Mock-infected mice treated with the inhibitor showed 100% survival and no signs of clinical disease ( Figure 9A ) . Interestingly , in the case of mice treated with the p38 MAPK inhibitor and infected with the parental virus , survival increased to 80% as compared with non-treated mice , with a mortality of 100% . SB203580 inhibits p38 MAPK catalytic activity by binding to the ATP-binding pocket , blocking the activation of several proteins regulated by the p38 MAPK pathway , including the heat-shock protein 27 ( HSP27 ) [47] , but does not inhibit phosphorylation of p38 MAPK by upstream kinases [48] . Therefore , to analyze whether SB203580 actually reduced p38 MAPK activity in lung tissue of virus-infected mice , the activation of HSP27 was studied by Western blot analysis at 2 dpi , using a phospho-HSP27 ( p-HSP27 ) specific antibody to detect the active form , and an antibody specifically recognizing the total endogenous HSP27 . Actin served as loading control . Results showed that the levels of active HSP27 were significantly reduced in the lungs of mice treated with SB203580 and infected with the parental virus , compared to those that were infected and non-treated ( Figures 9B and 9C ) , indicating that p38 MAPK activity was diminished by SB203580 . These results indicated that p38 MAPK activation is an important factor for SARS-CoV-induced disease .
Cellular factors containing PDZ domains participate in a complex network of protein-protein interactions that modulate many diverse biological processes such as cell polarity , cell-cell interactions , control of proliferation , migration , immune cell recognition and signal transduction pathways [28] , [49] , [50] . Alterations of these highly regulated processes can lead to important disorders , including several types of cancer [51] . Viruses have evolved proteins containing PBM to exploit these cellular networks for their own benefit , enhancing viral replication , dissemination in the host or pathogenicity [28] . Previously , we have shown that deletion of SARS-CoV E gene leads to an attenuated virus [18] , [19] , [21] . In this study , we focused on the contributions of the E protein PBM , a motif that actively participates in protein-protein interactions with host factors [23] , to the virulence of SARS-CoV . To this end , different mutant viruses containing altered or deleted E protein PBM sequences were generated using an infectious cDNA clone encoding a SARS-CoV adapted to efficiently grow in mice . Mutant viruses , with or without E protein PBM , grew in Vero E6 and DBT-mACE2 cells with titers similar to those reached by the parental virus . This result indicated that SARS-CoV E protein PBM is not essential for virus replication in cell culture . Interestingly , recombinant SARS-CoVs lacking E protein PBM were attenuated in vivo , causing minimal lung damage , and no mortality in infected mice . In contrast , viruses with functional E protein PBM were highly pathogenic causing 100% mortality and inducing profuse areas of damage in the lung , indicating that E protein PBM is a determinant of pathogenicity . SARS-CoV infection induces an exacerbated immune response that potentiates both epithelial and endothelial damage within the lungs , finally leading to edema accumulation , the ultimate cause of acute lung injury ( ALI ) and acute respiratory distress syndrome ( ARDS ) [52]–[54] . Both the enhanced immune response , which leads to cellular infiltration , and edema accumulation leading to pulmonary failure and death , occur when conventional mice are infected with a mouse adapted SARS-CoV [55] . In SARS-CoV-infected patients and animal models , it has been shown that the observed pathology is associated with an exacerbated inflammatory response , linked to elevated levels of pro-inflammatory cytokines [54] , [56] , [57] . To understand the mechanisms leading to attenuation of viruses lacking E protein PBM , differential host gene expression in mice infected with recombinant viruses with or without an E protein PBM was analyzed using microarray analysis . The expression of genes involved in the innate immune response was significantly reduced in mice infected with SARS-CoV lacking E protein PBM , suggesting an important role of the PBM in the uncontrolled immune response triggered during SARS-CoV infection . To further understand the molecular basis of the exacerbated immune response induced during SARS-CoV infection in the presence of SARS-CoV E protein PBM , host factors interacting with this motif were identified using a yeast two-hybrid system . One of the most prominent interactions was with the cellular protein syntenin , an important scaffolding protein containing two PDZ domains . PALS1 , a protein previously associated to E protein in a similar study [23] was not identified here , probably due to the use of different cDNA libraries . In this work a human lung cDNA library was chosen to further mimic SARS-CoV infection , whereas PALS1 was identified using a human placenta cDNA collection . Coprecipitation analysis revealed that the interaction of syntenin and SARS-CoV E protein was specifically mediated by a functional PBM located in the last 4 amino acid of the E protein , which most likely associates with syntenin PDZ domains , as E protein lacking PBM did not coprecipitate with syntenin . Syntenin may participate in the activation of p38 MAPK [41] , a crucial protein involved in the activation of a variety of transcription factors , controlling the expression of genes encoding inflammatory cytokines [42] , [43] . Indeed , we have shown an increased expression of inflammatory cytokines in vivo during the infection with recombinant SARS-CoVs containing E protein PBM , as compared with viruses lacking this motif , by using microarrays and RT-qPCR assays . Therefore , our results strongly suggest that the interaction of E protein with syntenin induced p38 MAPK activation leading to the inflammatory response observed after SARS-CoV infection . Interestingly , we have shown that in fact , p38 MAPK activation was significantly reduced in mice infected with viruses lacking E protein PBM as compared with mice infected with viruses containing a functional E protein PBM . This reduction in p38 MAPK activation correlated with the acquisition of an attenuated phenotype in SARS-CoV lacking the PBM . In addition , administration of one p38 MAPK inhibitor increased mouse survival after infection with the parental virus . The hypothesis that p38 MAPK is involved in CoV virulence was also postulated when patients with SARS showed augmented p38 MAPK activation [58] . It is clear that multiple host proteins and pathways may be activated through PDZ interactions with SARS-CoV E protein PBM . Interestingly , syntenin and PALS1 interaction with E protein might provide different and even complementary contributions to SARS-CoV pathogenesis , as syntenin has been described in this manuscript that activates p38 MAPK triggering an inflammatory response , whereas PALS1 affects the disruption of the lung epithelium in SARS patients [23] . In agreement with these results , proteins containing PBMs encoded by different viruses such as influenza A virus , tick-borne encephalitis virus ( TBEV ) and human papillomavirus ( HPV ) also enhanced virus pathogenesis by interacting with cellular proteins containing PDZ domains , by altering processes such as apoptosis , cell polarity or innate immune responses [28] . Our results are most likely of relevance to other coronaviruses , as the PBM is a highly conserved domain among most coronavirus E proteins , including the highly pathogenic MERS-CoV ( Figure S1 ) . Interestingly , although we demonstrated that E protein PBM was not essential for virus production after SARS-CoV infection , not all coronaviruses tolerate amino acid changes in the carboxy-terminus of E protein , as alanine substitutions in MHV E protein carboxy-terminal residues were apparently lethal , since no virus was recovered [59] . Nevertheless , it has not been excluded whether other viral proteins different from E provide alternative PBMs in the cases where the PBM does not seem to be essential . In summary , we have shown that SARS-CoV E protein contains a functional PBM that contributes to viral pathogenesis and interacts with several cellular proteins , including syntenin . Our studies strongly suggest a causal relationship between E protein-syntenin interaction and p38 MAPK activation , leading to an increase in inflammatory cytokines expression during infection . These data identify syntenin as a potential therapeutic target to reduce the exacerbated immune response induced during SARS-CoV infection . Targeted therapies that temporarily inactivate syntenin or p38 MAPK activation during acute infection may provide a rational approach to improve the prognosis in SARS-CoV patients . In fact , we have shown that inhibition of p38 MAPK increased mice survival after infection with SARS-CoV . Accordingly to our results , in vivo inhibition of p38 MAPK diminished influenza virus induced cytokine expression , protecting mice from lethal disease [60] . In the future , a search for the presence of additional PDZ targets located in alternative cellular proteins , and the relevance of E protein PBM present in other coronaviruses will be pursued to better understand the influence of PDZ and PBM motifs in virus pathogenicity and in the immune responses to virus infection .
Animal experimental protocols were approved by the Ethical Committee of The Center for Animal Health Research ( CISA-INIA ) ( permit numbers: 2011-009 and 2011-09 ) in strict accordance with Spanish National Royal Decree ( RD 1201/2005 ) and international EU guidelines 2010/63/UE about protection of animals used for experimentation and other scientific purposes and Spanish National law 32/2007 about animal welfare in their exploitation , transport and sacrifice and also in accordance with the Royal Decree ( RD 1201/2005 ) . Infected mice were housed in a ventilated rack ( Allentown , NJ ) . African Green monkey kidney-derived Vero E6 cells were kindly provided by Eric Snijder ( Medical Center , University of Leiden , The Netherlands ) . The delayed brain tumor ( DBT ) cells expressing mACE2 receptor ( DBT-mACE2 ) were generated in our laboratory [33] . Virus titrations were performed in Vero E6 cells as previously described [18] . The pcDNA3-E plasmid encoding SARS-CoV E protein was used as previously described [15] . The N-terminal HA-tagged syntenin expression plasmid in the backbone of pMT2-HA vector was kindly provided by P . J . Coffer ( University Medical Center , Utrecht , The Netherlands ) [61] . 8 week-old specific-pathogen-free BALB/c Ola Hsd mice females were purchased from Harlan Laboratories . BALB/c mice were infected at the age of 16 weeks with 100 , 000 plaque forming units ( pfu ) . Subconfluent monolayers ( 90% confluency ) of Vero E6 and DBT-mACE2 were infected at an MOI of 0 . 05 with wt , ΔE , ΔPBM , mutPBM and potPBM . Culture supernatants were collected at different hpi and virus titers were determined as previously described [18] . Lungs harvested for virus titers were weighed and homogenized using gentleMACS Dissociator ( miltenyibiotec ) . Virus titers were determined by plaque assay on Vero cells as previously described [18] . Mice were sacrificed at 2 and 4 dpi . Lungs were removed , fixed in zinc formalin and paraffin embedded . Histopathological examinations were performed on hematoxylin-eosin stained sections . Cell lysates were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) , transferred to a nitrocellulose membrane by wet immunotransfer and processed for Western blotting . The blots were probed with monoclonal antibodies specific for HA tag ( dilution 1∶10 , 000; Sigma ) , p38 MAPK ( dilution 1∶500; Cell Signaling ) , phospho-p38 MAPK ( dilution 1∶500; Cell Signaling ) , syntenin ( dilution 1∶1000; Abcam ) , phospho-HSP27 ( dilution 1∶1000 , Cell Signaling ) and actin ( dilution 1∶10000; Abcam ) or polyclonal antibodies against E ( dilution 1∶1000 ) , HSP27 ( dilution 1∶1000 , Cell Signaling ) and histone H3 ( dilution 1∶5000; Active Motif ) . A polyclonal antibody recognizing the carboxy-terminal domain of SARS-CoV E protein except the PBM was generated by Biogenes ( Germany ) using a synthetic peptide corresponding to the 49–64 residues of SARS-CoV E protein ( VSLVKPTVYVYSRVKN ) as previously described [15] . Bound antibodies were detected with horseradish peroxidase-conjugated goat anti-rabbit or anti-mouse antibodies ( dilution 1∶30 , 000; Cappel ) and the Immobilon Western chemiluminescent substrate ( Millipore ) . Vero E6 cells were grown to 90% confluency on glass coverslips and infected with the parental virus at an MOI of 0 . 3 . Alternatively , Vero E6 cells were grown to 70% confluency in 1 cm2 wells and transfected with 1 µg of DNA using 1 µl of Lipofectamine 2000 ( Invitrogen ) according to the manufacturer's instructions . At 24 hours post infection ( hpi ) or post transfection ( hpt ) , cells were fixed as previously described [15] . Primary antibody incubations were performed in PBS containing 10% FBS and 0 . 2% saponin for 1 h 30 min at room temperature . Immunofluorescence was performed using monoclonal antibodies specific for E ( dilution 1∶3000 ) or N ( dilution 1∶500 ) proteins [15] , and polyclonal antibodies specific for syntenin ( dilution 1∶200 , Abcam ) . Coverslips were washed four times with PBS between primary and secondary antibody incubations . Alexa 488- or Alexa 546-conjugated antibodies specific for the different species ( dilution 1∶500 , Invitrogen ) were incubated for 45 min at room temperature in PBS containing 10% FBS and 0 . 2% saponin . Nuclei were stained using DAPI ( dilution 1∶200 , Sigma ) . Coverslips were mounted in ProLong Gold anti-fade reagent ( Invitrogen ) and examined on a Leica SP5 confocal microscope ( Leica Microsystems ) . Vero E6 were grown to 90% confluence and transfected with a N-terminal HA-tagged syntenin expression plasmid . 24 hours later , cells were infected with recombinant viruses at an MOI of 0 . 3 . At 24 hpi , cell extracts were collected as previously described [62] . For immunoprecipitation assays , monoclonal anti-HA agarose conjugate clone HA-7 ( Sigma ) was used following the manufacturer's instructions . Briefly , 75 µl of the anti-HA agarose conjugate was washed five times with PBS and then incubated with the cell extracts overnight on an orbital shaker at 4°C . The samples were washed four times with PBS and then immune complexes were eluted using 20 µl 2X SDS sample buffer and heating at 95°C for 3 minutes . For reciprocal immunoprecipitation assays Protein A/G Plate IP Kit ( Pierce ) was used following the manufacturer's instructions as previously described [62] using polyclonal anti-E antibody . Analysis of precipitate complexes was carried out by SDS-PAGE and Western blotting . Lung sections from infected animals were collected at 2 dpi and homogenized using gentleMACS Dissociator ( Miltenyibiotec ) . Then , total RNA was extracted using the RNeasy purification kit ( Qiagen ) . Reactions were performed at 37°C for 2 h using a High Capacity cDNA transcription kit ( Applied Biosystems ) using 100 ng of total RNA and random hexamer oligonucleotides . Cellular gene expression was analyzed using TaqMan gene expression assays ( Applied Biosystems ) specific for mouse genes ( Table 1 ) . Data representing the average of three independent experiments were acquired and analyzed as previously described [32] . All experiments and data analysis were MIQE compliant [63] . Bait cloning and yeast two-hybrid screening with the carboxy-terminal ( amino acids 36–76 ) domain of SARS-CoV E protein ( ECT ) as bait were performed by Hybrigenics ( France ) . ECT domain was cloned into the pB27 vector , enabling its fusion with the LexA binding domain . The bait construct was transformed into the L40ΔGAL4 yeast strain [64] and then mated with the Y187 yeast strain transformed by a random-primed human lung cDNA library containing 10 million independent fragments . In the screening , 80 . 3 million interactions were analyzed . After selection on medium lacking leucine , tryptophan , and histidine , 268 positive clones were picked . The corresponding prey fragments were subjected to PCR and sequencing . Sequences were then filtered , divided into contigs , and compared to the latest release of the GenBank database by using BLASTn ( NCBI ) . A predicted biological score ( PBS ) was attributed to assess the reliability of the interaction , as described earlier [38] . Lungs were removed from infected mice at 2 dpi and homogenized . Nuclear and cytoplasmic extracts from homogenized lungs were obtained using a nuclear extract kit ( Active Motif , Carlsbad , CA ) . Levels of total and phosphorylated p38 MAPK were analyzed by Western blot using specific antibodies and the cytoplasmic extracts . Total and activated p38 MAPK amounts were quantified by densitometric analysis using Quantity One , version 4 . 5 . 1 , software ( Bio-Rad ) . In each case , the levels of phosphorylated p38 MAPK were normalized to the levels of total p38 MAPK . Three different experiments and appropriate gel exposures were used in all cases with similar results . In addition , different exposures of the same experiment were analyzed to assure that data obtained were within linear range . At 2 days post infection , lungs from infected mice were collected and homogenized using the gentleMACS Dissociator ( Miltenyibiotec ) . Then , total RNA was extracted using the RNeasy purification kit ( Qiagen ) according to the manufacturer's instructions . Three biological replicates were independently hybridized for each transcriptomic comparison . Total RNA ( 200 ng ) was amplified using One Color Low Input Quick Amp Labeling Kit ( Agilent Technologies ) and purified with RNeasy Mini Kit ( Qiagen ) . Preparation of probes and hybridization was performed as described in One-Color Microarray Based Gene Expression Analysis Manual Ver . 6 . 5 , Agilent Technologies . Briefly , for each hybridization 600 ng of Cy3 probes were mixed and added to 5 ul of 10x Blocking Agent , 1 ul of 25x Fragmentation Buffer and Nuclease free water in a 25 µl reaction , incubated at 60°C for 30 minutes to fragment RNA and stopped with 25 ul of 2x Hybridization Buffer . The samples were placed on ice and quickly loaded onto arrays , hybridized at 65°C for 17 hours in a Hybridization oven rotator and then washed in GE wash buffer 1 at room temperature ( 1 minute ) and in GE Wash Buffer 2 at 37°C ( 1 minute ) . Arrays were dried by centrifugation at 2000 rpm for 2 minutes . Slides were Sure Print G3 Agilent 8×60K Mouse ( G4852A-028005 ) Images were captured with an Agilent Microarray Scanner and spots quantified using Feature Extraction Software ( Agilent Technologies ) . Background correction and normalization of expression data were performed using LIMMA [65] . Linear model methods were used for determining differentially expressed genes . Each probe was tested for changes in expression over replicates by using an empirical Bayes moderated t-statistic [66] . To control the false discovery rate ( FDR ) , defined as the expected proportion of false positives among the significant tests , p-values were corrected by using the method of Benjamini and Hochberg [66] , [67] . The expected false discovery rate was controlled to be less than 5% ( FDR<0 . 05 ) . Genes were considered differentially expressed when the FDR were <0 . 01 . In addition , only genes with a fold change of >2 or of <−2 were considered for further analysis . The mouse-adapted ( MA15 ) [31] , parental virus ( wt ) and a virus lacking E gene ( ΔE ) were rescued from infectious cDNA clones generated in our laboratory [21] . The pBAC-SARS-CoV-E-PBM mutant plasmids encoding recombinant SARS-CoVs expressing E genes with deleted or mutated PBMs were constructed from a previously generated infectious cDNA clone ( plasmid pBAC-SARS-CoV-ΔE-MA15 ) [21] . Deletion of the 9 most carboxy-terminal amino acids ( LNS--------- ) or mutations ( LNSSAGAPALAV ) in SARS-CoV-E-ΔPBM and SARS-CoV-E-potPBM , respectively ( Figure 1 ) , were introduced by overlap extension PCR using the pBAC-SARS-CoV-ΔE-MA15 as a template and specific primers ( Table 2 ) . A fragment representing the nucleotides containing the mutations ( LNSSEGVPGGGG ) to generate SARS-CoV-E-mutPBM was chemically synthesized ( BioBasic Inc ) . The final PCR products and synthesis fragments were digested with enzymes BamHI and MfeI and cloned into the intermediate plasmid psl1190+BamHI/SacII-SARS-CoV to generate the plasmids psl1190-E-ΔPBM , psl1190-E-mutPBM and psl1190-E-potPBM . The plasmid psl1190+BamHI/SacII SARS-CoV contains a fragment corresponding to nucleotides 26045 to 30091 of the SARS-CoV infectious cDNA clone [68] engineered into plasmid psl1190 ( Pharmacia ) . These constructs were cloned in the infectious pBAC-SARS-CoV-ΔE-MA15 with the enzymes BamHI and SacII . All constructs were generated carrying a duplication of the final last nucleotides ( 208–231 ) of E gene after the stop codon of the mutated E proteins in order to avoid altering the transcription regulatory sequence ( TRS ) of membrane ( M ) gene , which overlaps with the end of E gene [69] . All viruses were rescued from infectious cDNA clones as previously described [68] . Vero E6 cells were transfected following a reverse transfection protocol . Briefly , for each well of a 24-well plate , 5×104 cells were incubated in suspension with 25 nM of Silencer Select siRNAs ( Ambion ) targeting syntenin ( siRNA 1: s12641 and siRNA 2: s224582 ) and 2 µl of siPORT amine ( Ambion ) diluted in 50 µl of Opti-MEM I reduced serum medium ( GibcoBRL-Invitrogen ) , following the manufacture's instructions . As a negative control , an irrelevant validated siRNA ( Ambion , reference 4390843 ) was transfected . Cells were plated onto each well using DMEM with 5% heat-inactivated FBS , incubated at 37°C for 24 h . The cells were retransfected with 25 nM siRNAs at 24 h after the first transfection , and infected with the parental virus at 48 h after second transfection . At 24 hpi , total RNA , proteins and cells supernatants were collected for further analysis . Syntenin gene expression was quantified by qRT-PCR . Total RNA was prepared with an RNeasy kit ( Qiagen ) , according to the manufacturer's instructions . Reactions were performed at 37°C for 2 h using a High Capacity cDNA transcription kit ( Applied Biosystems ) using 100 ng of total RNA and random hexamer oligonucleotides . Syntenin gene expression was analyzed using TaqMan gene expression assays ( Applied Biosystems ) specific for human gene ( Table 1 ) . Data representing the average of three independent experiments were acquired and analyzed as previously described [32] . All experiments and data analysis were MIQE compliant [63] . 16-week-old BALB/c mice were infected intranasally with 100 , 000 pfu of wt virus . At 4 hpi and every 12 h thereafter , from days 1 to 8 , mice were treated intraperitoneally with SB203580 ( Millipore ) at 6 mg/kg of body weight/day with vehicle ( PBS containing 2% dimethyl sulfoxide [DMSO] ) . Survival was analyzed in three independent experiments with 5 mice per group . To analyze p38 MAPK inhibition , lungs were removed and homogenized from mice at 2 dpi . Levels of total and phosphorylated HSP27 were analyzed by Western blot using specific antibodies and the lung extracts . Total and activated HSP27 amounts were quantified by densitometric analysis using Quantity One , version 4 . 5 . 1 , software ( Bio-Rad ) . In each case , the levels of phosphorylated HSP27 were normalized to the levels of total HSP27 . Three different experiments and appropriate gel exposures were used in all cases with similar results . In addition , different exposures of the same experiment were analyzed to assure that data obtained were within linear range . Vero E6 were washed twice with PBS , scraped off and pelleted by low-speed centrifugation at 2000 rpm for 2 minutes in a bench-top centrifuge . The supernatant was removed and cell pellets were lysed by repetitive pipetting with a micropipette in ice-cold lysis buffer containing 150 mM NaCl , 3 mM MgCl2 , 20 mM Tris/HCl ( pH 7 . 5 ) , 2 mM DTT and 0 . 5% NP-40 . The lysate was incubated during 5 minutes in ice and centrifuged in a bench-top centrifuge at 3000 rpm for 2 minutes . Supernatant and pellet were saved as the cytosolic and nuclear fraction , respectively . Hematoxylin and eosin-stained lung sections were assessed using the scoring system described in the figure legends according to previously described procedures [34] . Three animals for each time point were analyzed . The UniProt ( http://www . uniprot . org/ ) accession numbers for genes and proteins discussed in this paper are: SARS-CoV E protein , P59637; human p38 MAPK , Q16539; mouse p38 MAPK , P47811; syntenin , O00560; human ACE2 , Q9BYF1; mouse ACE2 , Q8R0I0; PALS1 , Q8N3R9; SAA2 , P05367; CCL3 , P10855; CXCL1 , P12850; CXCL5 , P50228; CALCA , P70160; SAA1 , P05366; CXCL10 , P17515; CCL2 , P10148; IL1B , P10749; ORM1 , Q60590; IL6 , P08505; CCL4 , P14097; CXCL9 , P18340; 18S , O35130; human actin , P60709; mouse actin , P60710; histone H3 , P84243; SARS-CoV N protein , P59595; HSP27 , P14602 . | SARS-CoV caused a worldwide epidemic infecting 8000 people with a mortality of about 10% . A recombinant SARS-CoV lacking the E protein was attenuated in vivo . The E protein contains a PDZ-binding motif ( PBM ) , a domain potentially involved in the interaction with more than 400 cellular proteins , which highlights its relevance in modulating host-cell behavior . To analyze the contributions of this motif to virulence , recombinant viruses with or without E protein PBM were generated . Recombinant SARS-CoVs lacking E protein PBM caused minimal lung damage and were attenuated , in contrast to viruses containing this motif , indicating that E protein PBM is a virulence determinant . E protein PBM induces the deleterious exacerbated immune response triggered during SARS-CoV infection , and interacts with the cellular protein syntenin , as demonstrated using proteomic analyses . Interestingly , syntenin redistributed from nucleus to cytoplasm during SARS-CoV infection , activating p38 MAPK and triggering the overexpression of inflammatory cytokines . Furthermore , silencing of syntenin using siRNAs led to a decrease in p38 MAPK activation . In addition , administration of a p38 MAPK inhibitor led to an increase in mice survival after SARS-CoV infection . These results indicate that syntenin and p38 MAPK are potential therapeutic targets to reduce the exacerbated immune response during SARS-CoV infection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"emerging",
"infectious",
"diseases",
"infectious",
"diseases",
"inflammation",
"pathogenesis",
"immune",
"response",
"immune",
"system",
"immunopathology",
"clinical",
"immunology",
"immunity",
"host-pathogen",
"interactions",
"virology",
"virulence",
"factors",
"biology",
"and",
"life",
"sciences"
]
| 2014 | The PDZ-Binding Motif of Severe Acute Respiratory Syndrome Coronavirus Envelope Protein Is a Determinant of Viral Pathogenesis |
Spinal muscular atrophy is a severe neurogenic disease that is caused by mutations in the human survival motor neuron 1 ( SMN1 ) gene . SMN protein is required for the assembly of small nuclear ribonucleoproteins and a dramatic reduction of the protein leads to cell death . It is currently unknown how the reduction of this ubiquitously essential protein can lead to tissue-specific abnormalities . In addition , it is still not known whether the disease is caused by developmental or degenerative defects . Using the Drosophila system , we show that SMN is enriched in postembryonic neuroblasts and forms a concentration gradient in the differentiating progeny . In addition to the developing Drosophila larval CNS , Drosophila larval and adult testes have a striking SMN gradient . When SMN is reduced in postembryonic neuroblasts using MARCM clonal analysis , cell proliferation and clone formation defects occur . These SMN mutant neuroblasts fail to correctly localise Miranda and have reduced levels of snRNAs . When SMN is removed , germline stem cells are lost more frequently . We also show that changes in SMN levels can disrupt the correct timing of cell differentiation . We conclude that highly regulated SMN levels are essential to drive timely cell proliferation and cell differentiation .
Proximal spinal muscular atrophy ( SMA ) is characterised by the loss of the α-motor neurons in the anterior horns of the spinal cord , leading to progressive paralysis , muscle wasting , and in the most severe cases , death . SMA , an autosomal recessive disease , is the most common genetic form of infant mortality with an incidence of 1 in 10 , 000 live births [1] . It is caused by mutations or deletions in the survival motor neuron 1 ( SMN1 ) gene which , together with a paralogue SMN2 , lies within an inverted repeat on human chromosome 5q13 [2] , [3] . Due to altered splicing efficiency SMN2 produces levels of SMN protein that are too low to maintain healthy motor neurons [4] , [5] , [6] . SMN is a ubiquitously expressed protein and functions within a large multiprotein complex that recruits and assembles small nuclear ribonucleoproteins ( snRNPs ) . snRNPs are components of the macromolecular spliceosome that catalyses the splicing of pre-mRNAs [7] . Additional functions that have also been attributed to SMN include the processing of additional RNA subclasses and mRNA processing and transport in axons [8] , [9] . However , how the reduction of SMN protein leads to a neuronal specific disease remains elusive [10] . SMN protein is highly expressed in the early mouse , zebrafish and Drosophila embryos [11] , [12] , [13] . In whole mouse tissues , snRNP-associated SMN activity is down-regulated upon differentiation [11] . Developmental defects have been observed in a number of models , in particular zebrafish , which display early axonal branching defects [14] . However , it is still unknown which cell populations within the developing tissues have higher SMN levels and how the protein is regulated on an individual cell level . To understand the role of SMN in disease it is therefore important to understand 1 ) the unique vulnerability of motor neurons to the deficiency of this ‘housekeeping gene’ 2 ) why a monogenic deficiency causes a wide spectrum of phenotypic severity and 3 ) whether defects in SMA are determined early in development or related to degeneration later in life [15] . This study uses the tractability of the Drosophila system to uncover how developing tissues respond to SMN level changes . Here we report observations of SMN expression in two well-defined tissues in Drosophila: the larval CNS , and the male germline . We found that in both tissues the stem cells display the highest levels of SMN . SMN levels then decrease in a consistent gradient as cells differentiate into mature neurons and sperm . If SMN is removed from stem cells , division is slower in the CNS and stem cell loss is more frequent in the testis . SMN mutant neuroblasts have abnormally localised Miranda , which is an adaptor protein that binds and facilitates the basal anchoring of prospero mRNA in neuroblasts . Proliferation defects also correlate with snRNP reduction in the developing CNS and in the germline . In the developing testis , we show that contraction of the SMN gradient leads to premature differentiation , while its expansion can repress differentiation . Taking these results together , we conclude that the tight regulation of SMN expression on a cellular level is important for stem cell division , proliferation and daughter cell differentiation .
We analysed the Drosophila loss of function alleles smnA ( smn73Ao ) and smnB which survive on maternally contributed wild-type SMN supplied from the heterozygous mother . smnA and smnB larvae develop motor defects and die at 2nd and 3rd instars , respectively [12] , [16] . Prior to the onset of motor defects , both SMN mutants displayed CNS growth defects ( Figure 1A , wild-type; 1B , less severe smnB only ) . As flies are holometabolic insects that undergo metamorphosis , their larval CNS comprises of regions of both fully differentiated and developing neurons for the respective larval and adult stages [17] . During larval life , postembryonic neuroblasts ( pNBs ) exit quiescence , enlarge and divide to generate the neurons , including motor neurons , required in the adult fly . These neurons remain in an immature state and can be observed in the brain lobes and the thoracic and abdominal ganglion . Both smnA and smnB mutant CNS were reduced in size when compared to wild-type at day 4 and 5 . smnA CNS did not increase in size after this stage and the larvae die soon after the day 4 measurement . As Shpargel and colleagues previously described , smnB mutants can survive up to and beyond 8 days where they die as 3rd instar larvae or as pseudopupae [16] . The size of smnB CNS at day 8 failed to reach that of CNS from wild-type larvae at day 5 . Although a ubiquitously expressed protein , SMN has been shown to be regulated during cell differentiation . To understand how SMN may control the generation of new neurons , we analysed SMN levels in the ventral ganglion of the developing larval CNS ( Figure 2 ) . In the early 1st instar larval CNS , SMN staining was ubiquitous and localised in punctate bodies . During the late 1st and 2nd instar stages SMN levels increased in cells that correspond to the quiescent pNBs ( Figure 2E–2G ) . SMN enrichment coincided with the expression of Grainyhead ( Grh ) , a transcription factor and pNB marker , [18] and prior to the detection of Miranda ( Figure 2 ) . SMN accumulation increased as the pNBs enlarged with the highest expression of SMN protein found in the cytoplasm of 2nd and 3rd instar dividing pNBs ( Figure 2A–2C; 3rd instar pNBs ) . Each pNB divides asymmetrically producing a large cell , which retains neuroblast identity , and a smaller cell termed the ganglion mother cell ( GMC ) , which divides terminally into two postmitotic progeny . The intensity of SMN expression in the adjacent GMC was slightly lower . Levels in the cytoplasm then decreased in a gradient through the daughter cells until it resided to a basal level in the differentiated neurons and glia . In the Drosophila larval brain both type I ( ID and IA ) and type II neuroblasts generate progeny through different intermediary precursor cells [19] , [20] . We have found SMN is enriched in type ID ( thoracic and brain lobe ) , type IA and all the Miranda staining neuroblasts of the brain lobes ( Figure S1 ) . SMN functions within a complex with a group of proteins called the Gemins . We have also found that Gemin5 has a comparable pattern in the CNS ( Figure S2 ) . Miranda is a cargo protein that forms a messenger ribonucleoprotein ( mRNP ) complex with prospero , a transcription factor that drives daughter cell differentiation [21] . Before division , Miranda protein localises to the apical membrane of the neuroblast directing the basal localisation of prospero mRNA [22] . These proteins are arranged on the basal membrane during late prophase and metaphase and become segregated into the GMC upon cytokinesis . To further understand the function of SMN in pNBs , we looked to see if Miranda localisation was affected ( Figure 3 ) . In wild-type cells Miranda was asymmetrically localised as a crescent on the neuroblast membrane parallel to the metaphase chromosomes ( Figure 3A , 3B ) . The metaphase chromosomes were labelled with phosphorylated histone H3 ( pH 3 ) . pH 3 is a marker that is expressed at M phase . Phosphorylation of histone H3 on serine-10 promotes the condensation of chromatin , an event tightly linked to the entry into mitosis . Both smnA ( Figure 3C , 3D ) and smnB ( Figure 3E , 3F ) neuroblasts displayed defective Miranda localisation . In 23% of smnB neuroblasts ( n = 91 ) and 71% of smnA neuroblasts ( n = 42 ) , Miranda did not correctly localise in a crescent during metaphase , and could be seen to be diffuse or punctate in the cytoplasm , or in a band that was not localised correctly in relation to the metaphase chromosomes ( Figure 3G ) . We used the mosaic analysis with a repressible cell marker ( MARCM ) system to visualize pNB clones lacking SMN with the positive marker GFP [23] . When double-stained with antibodies against SMN and GFP , smn MARCM clones ( GFP positive ) had a low SMN signal ( Figure 4A , 4B ) , in contrary to their neighbouring GFP-negative cells . In wild-type clones SMN enrichment was still present ( Figure S3 ) . The GFP clones represent pockets of immature neurons that will become the motor neurons and glial cells of the adult fly . To analyse how SMN reduction intrinsically affects the thoracic type IID larval neuroblasts , we heat-shocked wild-type and smnA MARCM crosses and analysed clones at 65- , 82- and 96-hr post hatching . Quantitative analysis shows smnA mutant pNB clones were significantly smaller than control clones derived from wild-type pNBs ( Figure 4C , 4E ) . GMCs separated from the pNB in an inconsistent pattern generating mutant clones that were abnormally arranged ( Figure 4E ) . Many of the smnA clones resided on the ventral surface and often daughter cells were seen adjacent to but not completely part of the pockets of cells ( Figure 4B and 4E ) . Clones at the 96-hr stage were also stained for pH 3 . The number of wild-type and smn clones with pNBs and GMCs positive for pH 3 were counted and displayed as a percentage of the total number of clones from each type ( Figure 4F–4H ) . Wild-type control clones were stained with pH 3 , a mitotic marker , in 48% of the clones . In contrast only 24% of smnA clones were shown to be in M phase at that time . This suggests that the smn mutant clones are in M phase less often and therefore dividing less than wild-type pNBs . Both smnA and smnB mutant larvae had a reduction of pH 3 in the thoracic ganglion ( Figure S4 ) . To see if stem cells within the clones were lost we stained 96-hr old clones for the pNB marker Grh . Almost all ( 98% ) of the smnA mutant clones contained a large Grh-positive cell suggesting stem cell loss had not occurred at this stage ( Figure 4I ) . smnA and smnB larvae were also stained for activated caspase-3 to test for apoptosis , and Hsp70 to test for stress . In all cases these markers were not present in the pNBs or GMCs in the thoracic ganglion of the mutants , suggesting the reduction in proliferation was not due to pNB death ( Figure S5 ) . SMN , in complex with a set of proteins called the Gemins , promotes the assembly of uridine-rich snRNPs which are components of the spliceosome . snRNPs consist of an Sm protein ring and a number of uridine-rich snRNAs that include U2 and U5 . In Drosophila , a minimal SMN complex consisting of SMN , Gemin2 ( SIP2 ) , Gemin3 and Gemin5 ( Rig ) has been reported [16] , [24] , [25] , [26] . To understand if SMN reduction in the dividing neuroblasts affects U snRNAs , U2 and U5 levels were tested using in situ hybridisation in smnA mutant clones ( Figure 5 ) [27] , [28] . It was previously reported that there are no gross changes in snRNA levels in smnA and smnB mutant larvae [29] . However in this study both U2 and U5 levels were reduced in smnA mutant MARCM clones suggesting snRNAs in the developing Drosophila neurons may be particularly sensitive to SMN reduction ( Figure 5C , 5D , 5G , 5H ) . U2 and U5 were observed in a consistent pattern in both the surrounding wild-type cells in the smnA MARCM clones , and in the wild-type MARCM clones ( Figure 5 ) . Drosophila testes offer a tractable system to study stem cell maintenance and cell differentiation . In addition , Drosophila testes have the highest number of alternative splicing events and exhibit prominent changes in the expression of snRNPs and splicing factors as sperm develop [30] . In addition to the larval CNS , SMN protein formed a striking gradient in Drosophila testes ( Figure 6 ) . In adult testes , 8–12 germline stem cells ( GSCs ) surround the hub cells which serve as the somatic niche . Each GSC divides into two daughter cells . One daughter cell remains as a stem cell , while another daughter cell differentiates into a gonialblast ( GB ) . GBs divide four times to produce 2- , 4- , 8- and 16-cell cysts ( Figure 6A ) [31] . Each spermatogonial cell in a 16-cell cyst will undergo meiosis giving rise to 4 spermatids . SMN staining was virtually undetectable in the hub cells , whilst it was enriched in GSCs . SMN levels remained very high in GBs and spermatogonia with predominant punctate structures resembling U bodies . SMN levels then decreased dramatically in spermatocytes ( Figure 6B–6G ) . SMN formed a clear gradient from GSCs to their differentiated progeny in adult testes ( Figure 6H ) . Stem cell loss was also observed in testis mitotic clones which were studied over an 11-day period ( Figure 7 ) . In contrast to MARCM , smn mutant cells are GFP negative when using the mitotic clonal system . Clones that contained GFP-negative gonialblasts and primary spermatocytes , but no GFP-negative GSCs , were regarded as having lost the GSCs that would have previously generated the GFP negative cells observed ( Figure 7A , 7B ) . This enables the identification of clones where the stem cell is lost . The number of smnA mitotic clones that lacked a GFP-negative stem cell was considerably higher than the wild-type clones in this study . This difference became more pronounced at day 11 indicating SMN is essential for the survival of GSCs in testes . Quantitative analysis showed that the longevity of male GSCs generating smnA clones was greatly reduced ( Figure 7C ) . To test whether snRNP levels were also altered in GSC clones , we analysed the mitotic clones with a combination of immunostaining and fluorescence in situ hybridisation . In smnA mutant clones U2 levels were reduced ( Figure 7D–7G ) . These results suggest that SMN is required for both the proliferation and survival of GSCs , and the maintenance of snRNA levels in stem cell clones . Our results have shown that SMN enrichment in larval neuroblasts and male GSCs is required for proper proliferation , snRNA levels and Miranda localisation . In addition , SMN levels are strongly downregulated upon differentiation . Using daughterless-GAL4 ( da-GAL4 ) and a functional SMN-YFP transgene , we looked to see how ubiquitous overexpression of SMN affects development . Flies with transgenic expression of SMN using da-GAL4 were viable but exhibited numerous growth defects . To understand how SMN overexpression affects larval CNS growth we measured the length of the ventral ganglion along the ventral midline ( Figure 8 ) . Embryos from each genotype were collected over a 1-hr laying period and grown at 25°C with comparable food amounts and population densities . The CNS from da-GAL4; SMN-YFP larvae increased in size quicker than that from wild-type animals ( Figure 8A , 8B ) . Upon pupation the pupae cases of da-GAL4; SMN-YFP also failed to contract fully and appeared elongated ( Figure 8D , 8E ) . However , these pupae were viable . In addition , da-GAL4; SMN-YFP larvae had irregular pupation patterns with many larvae pupated earlier than the da-GAL4 controls ( Figure 8F ) . We next wanted to understand how altering SMN levels can affect the downstream progeny of stem cells . We used Drosophila larval testis as a model since the SMN expression gradient is the most striking in this system ( Figure 9 ) . The 3rd instar larval testis is an oval-shaped organ with bands of distinct cell types residing along the apical-terminal axis . GSCs , GBs and cysts occupy the apical fifth , while terminal cells occupy the terminal fifth . The central three fifths of the testis consist predominantly of large spermatocytes which can be identified with bright coilin staining ( Figure 9B ) [32] . This pattern is consistently observed in wild-type larvae and is compatible with the correct temporal development of mature sperm for the adult fly . Consistent with adult testes , SMN formed a gradient in larval testes with high levels of SMN in GSCs , GBs and decreased levels in cysts and spermatocytes ( Figure 9A ) . As developing cells migrate further away from the apical stem cells , the level of SMN protein decreased . Based on the striking inverse correlation of SMN concentration and cell differentiation , we hypothesised that SMN could affect the differentiation of specific cell types . To test this idea , we examined larval testes with overexpressed and low levels of SMN ( Figure 9C–9E ) . Embryos were collected over a 2-hr laying period and testes were dissected 4 days after hatching . In 3rd instar smnB testis , residual amounts of SMN were present due to low levels of maternal contribution ( Figure 9D ) . An SMN gradient was still present in smnB testes but was restricted to the apical tip . Levels then sharply declined until undetectable in cells distal to the apical end . Consistent with the shift of the SMN gradient , the terminal side of the spermatocyte band retracted to the apical terminus causing an increase in the number of differentiated cells . Mature sperm were present at this stage within this region ( Figure 9F , 9G ) . Groups of cells in this region also underwent apoptosis . Overall the mutant testes were smaller than same-stage wild-type controls . In contrast , overexpressing SMN-YFP that was driven by da-GAL4 induced a migration of the SMN gradient towards the terminal end ( Figure 9E ) . da-GAL4; SMN-YFP was expressed in the whole testis , however the most intense expression was in the primary spermatocye band . Consequently , SMN levels appeared very low around the hub and stem cells ( Figure 9E ) . This high expression caused the apical boundary of the spermatocyte region moves towards the terminal end . The size of da-GAL4; SMN-YFP testes could be at least 3 times larger than testes from wild-type larvae . To understand this phenotype further the morphology and development of sperm was analysed in adult testes . Similarly , overexpression of SMN caused an increased number of primary spermatocytes in adults , creating an enlarged tumour-like phenotype ( Figure 9H , 9I ) . In addition very few mature sperm were present in the adult suggesting that ectopic expression of SMN represses the differentiation of spermatogonia into sperm .
This study shows a high demand for SMN in Drosophila stem cells . In addition , we have identified a striking SMN concentration gradient , inversely proportional to the state of differentiation , in Drosophila larval CNS and testis . In Drosophila SMN mutant larvae , both the CNS and testis display growth defects which precede the previously reported motor defects and death . These larvae also fail to localise Miranda protein correctly at the basal membrane of the neuroblast . Clonal analysis indicates that SMN deficient stem cells have a reduced number of divisions and also generate cells with lower levels of U2 and U5 snRNPs . Overexpression of SMN alters the timing of CNS growth and disrupts the onset of pupariation and pupation . Using the male germline system , we show that prolonged SMN reduction leads to stem cell loss . Finally we find that ectopic SMN expression in cells along the SMN gradient leads to changes in the timing of cell differentiation . We therefore suggest that the fine-tuning of SMN levels throughout development can lead to complex developmental defects and reduce the capacity of stem cells to generate new cells in development . SMN levels have been reported to be extremely high in early development [11] . We show that SMN up-regulation occurs in neuroblasts prior to the initiation of their cell division , suggesting a distinct increase of SMN levels is required for new rounds of neurogenesis and local proliferation . Fewer immature neurons are generated in the thoracic ganglion of smn mutant MARCM clones . Provisional data has suggested there may be proliferation defects in the spinal cord of severe mouse models [33] . In addition , a recent study using the severe SMA mouse model has shown proliferation defects in the mouse hippocampus , a region associated with higher SMN levels [34] . Together these data suggest that , in part , the pathology observed in more severe forms of SMA may be caused by defects in tissue growth . Proteins involved in processes such as chromatin remodelling , histone generation and cell signalling have been identified as intrinsic factors for the maintenance of Drosophila stem cells . To the best of our knowledge , this is the first report of stem cell defects caused by the reduction of a protein involved in snRNP biogenesis . Although SMN is required in all cells , proper stem cell function requires a substantially higher level of SMN . This study also shows snRNP defects in Drosophila SMN mutant tissue . Previous studies in Drosophila have shown no gross changes in snRNP levels , including U2 and U5 , in lysates from whole smnA and smnB mutant larvae [29] . smnA MARCM neuroblast clones and male germline mitotic clones have reduced snRNP levels , suggesting snRNP assembly may be particularly sensitive to SMN reduction during CNS and germline development . SMN mutant neuroblasts have abnormal Miranda localisation . Miranda , an adaptor protein , forms a complex with the RNA binding protein Staufen which binds to prospero mRNA [35] , [36] . In addition to snRNPs , SMN protein has been implicated in the biogenesis of numerous RNP subclasses , including proteins involved in the transport and localisation of β-actin mRNA at the synapse . Whether Miranda mislocalisation is due to direct or indirect associations with SMN should be addressed . SMN mutant larvae have been previously shown to have synaptic defects which include enlarged and fewer boutons and a reduction in the number of GluR-IIA clusters – the neurotransmitter receptor at the Drosophila neuromuscular junction [12] , [24] . In addition , numerous developmental defects are observed including pupation and growth defects . Complementing this work , Drosophila Gemin5 a member of the Drosophila SMN-Gemin complex has been shown to interact with members of the ecdysone signalling pathway responsible for initiating pupation and growth [37] . Drosophila Gemin5 is also enriched in pNBs , in a pattern comparable to SMN ( Figure S5 ) . There is increasing evidence that suggests the Drosophila SMN complex plays an important role in pupation . Ubiquitous overexpression of SMN using da-GAL4 advances CNS development and causes premature entry into pupation . The ecdysone pathway has been identified to play an important part in the regulation of neuroblast division and neuronal differentiation during development [38] , [39] . How the Drosophila SMN complex plays a part in stem cell biology , and how the SMN complex interacts with specific signalling pathways should be the subject of further study . Larval and adult testes exhibit the most distinct SMN gradients in Drosophila tissues . Drosophila testes have a constant population of germline stem cells that start to divide in the late larval stages and produce sperm throughout life . The removal of SMN from male germline stem cells results in stem cell loss . In the smnB mutant testis , the reduction of SMN causes a contraction of the SMN gradient towards the apical stem cells . As SMN is lost from the primary spermatoctyes , more mature sperm are observed . Increasing SMN levels leads to an increase in primary spermatocytes and a reduction in mature sperm in the adult . This result is the first to demonstrate that high SMN levels in undifferentiated cells can repress differentiation in sperm development . Interestingly , along with the CNS , Drosophila testes have the highest number of alternative splicing events and the most differentially expressed splicing factors during development [30] . Understanding if differential expression of SMN in specific cell types controls a shift in splicing factors as cells switch from proliferation to differentiation will be the target of future study . A recent study has identified defects in gametogenesis and testis growth in mice lacking the Cajal body marker coilin , a binding partner of SMN [40] . The authors speculated that coilin may facilitate the fidelity and timing of RNP assembly in the cell and coilin loss may limit rapid and dynamic RNA processing . It will be important to understand how SMN and coilin genetically interact in stem cells and developing tissues . The Drosophila CNS and male germline offer two new tractable systems that can be used to study SMN biology in development and stem cells . It also offers a system to study how SMN , a protein associated with neuronal development , could cause SMA . Although SMA is classically a disease of the motor neuron , a severe reduction of SMN protein affects a wide spectrum of cells including stem cells . Consistent with this idea , symptoms in mild forms of SMA ( type III or IV ) [41] are predominately limited to motor neurons . However , patients with the most severe type ( type I ) , suffer from defects in multiple tissues including congenital heart defects , multiple contractures , bone fractures , respiratory insufficiency , or sensory neuronopathy [42] , [43] , [44] , [45] , [46] . Elucidating the differential requirements of SMN in individual cell types , and how their sensitivity to SMN loss can mediate the disease , can contribute to the understanding of the selectivity of SMA .
smnA ( smn73Ao ) and smnB , two independently generated smn mutant alleles were used in this study [12] , [47] . Both contain point mutations in a highly conserved carboxy-terminal domain of the SMN protein which harbours a self dimerization YXXG motif [12] . A stock with an extra copy of SMN ( SMN-YFP driven by daughterless-GAL4 ) was used for SMN overexpression experiment [28] . The transgene SMN-YFP is functional as it could rescue smn mutation [29] . y w , OrgR or da-GAL4 was used as the control strains . To generate MARCM clones , hsFLP; smnA FRT2A/TM6b or control hsFLP; + FRT2A/TM6b females were crossed to males from the MARCM driver stock C155 ELAV-GAL4 , UAS mCD8::GFP; P{TubP-Gal80} , FRT2A/TM6B [48] . Embryos were collected over 2 hrs periods and then heat shocked at 18 to 24 hrs after laying at 37°C for 1 hr . Larvae were dissected at 65 , 82 , or 96 hrs for analysis . smnA and wild-type control mitotic clones were generated using FLP-mediated mitotic recombination in adult testes [49] . smnA stem cells and progeny in mitotic clones are GFP negative . To generate stocks for clonal analysis , Ubiquitous-GFP/TM3 , Ser flies were crossed to w hsFLP; smnA FRT2A/TM3 , Ser and w hsFLP; + FRT2A/TM3 , Ser flies . One-day-old non-Ser flies were selected for heat shocking at 37°C for two 1-hr heat shocks . Flies were fed on fresh wet yeast every day . Testes were analysed at 3 or 11 days . Stem cell maintenance and relative division rate calculations were determined as previously described by Xie and Spradling [50] . Tissue was dissected in Grace's Insect Medium , fixed in 4% paraformaldehyde in PBS and washed in 1× PBS+0 . 3% ( v/v ) Triton X-100 , 0 . 5% normal goat or horse serum ( PBT ) . To analyse the SMN gradient , larval CNS , adult testes were subjected to overnight staining at room temperature using three rabbit anti-SMN antibodies ( Marcel van Den Heuvel , 1∶500; Spyros Artavanis-Tsakonas , 1∶2000; and Jianhua Zhou 1∶2000 ) [12] , [51] and one mouse anti–SMN ( Spyros Artavanis-Tsakonas , 1∶500 ) [51] . MARCM and mitotic clones were analysed using mouse anti-GFP ( Abcam , 1∶500 ) and rabbit anti-GFP ( Roche , 1∶500 ) . Rabbit anti-Miranda ( C . Doe , 1∶200 ) , mouse anti-Grainyhead ( S . Bray , 1∶5 ) , mouse anti-HTS ( 1∶20 ) , hsp70 ( 1∶100 , Santa Cruz ) , activate caspase-3 ( 1 200 Abcam ) , pH 3 ( 1∶200 , Upstate Biotechnology ) were also used . Alexa-Fluor conjugated secondary goat antibodies were used at a 1∶250 to 1∶1000 concentration . Samples were counterstained with nuclear-staining Hoechst 33342 ( 1∶500 ) prior to viewing with a Zeiss LSM 510 META confocal microscope . Images were processed using Adobe Illustrator . All figures in this paper were generated using rabbit anti SMN ( 1∶2000; gift from Jianhua Zhou ) . The in situ protocol and probes have been previously described [27] , [28] . Tissue was dissected in Grace's medium and fixed in 4% paraformaldehyde in 1XPBS for detecting RNA , for 10 minutes at room temperature . Tissue is then washed in by 100 µl in situ mix ( Formamide 20XSSC Heparin ( 5 mg/ml ) , yeast tRNA ( 50 mg/ml ) , Citric acid ( 0 . 5 M pH 6 . 0 ) DEPC H2O , 20% TWEEN 20 ) for 5 minutes . Fluorescent probes were then added and incubated for 1 hr at 37°C . | Spinal muscular atrophy is a debilitating disease that affects the motor nervous system . The disease is caused by the reduction of the protein survival motor neuron ( SMN ) , which is involved in the assembly of ubiquitous small nuclear ribonucleoproteins . As SMN is required in every cell , it is important to understand the differential functionality of the protein within developing tissues . In this paper , we identify stem cells as having the highest levels of SMN . The concentration of SMN then decreases in a declining gradient until it reaches its lowest level in differentiated cells . SMN reduction , using clonal analysis , slows stem cell division and can lead to stem cell loss . These defects correlate with a reduction in the U2 and U5 small nuclear RNAs and with the mislocalisation of Miranda protein in postembryonic neuroblasts . In addition , we show that the overexpression of SMN can change the timing of development and cell differentiation . This research highlights possible mechanisms explaining how SMN expression alterations may affect tissue development . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"animal",
"models",
"medicine",
"developmental",
"biology",
"drosophila",
"melanogaster",
"model",
"organisms",
"neural",
"stem",
"cells",
"stem",
"cells",
"neurological",
"disorders",
"neurology",
"biology",
"cell",
"differentiation",
"motor",
"neuron",
"diseases",
"adult",
"stem",
"cells"
]
| 2011 | Survival Motor Neuron Protein Regulates Stem Cell Division, Proliferation, and Differentiation in Drosophila |
The early host response to pathogens is mediated by several distinct pattern recognition receptors . Cytoplasmic RNA helicases including RIG-I and MDA5 have been shown to respond to viral RNA by inducing interferon ( IFN ) production . Previous in vitro studies have demonstrated a direct role for MDA5 in the response to members of the Picornaviridae , Flaviviridae and Caliciviridae virus families ( ( + ) ssRNA viruses ) but not to Paramyxoviridae or Orthomyxoviridae ( ( − ) ssRNA viruses ) . Contrary to these findings , we now show that MDA5 responds critically to infections caused by Paramyxoviridae in vivo . Using an established model of natural Sendai virus ( SeV ) infection , we demonstrate that MDA5−/− mice exhibit increased morbidity and mortality as well as severe histopathological changes in the lower airways in response to SeV . Moreover , analysis of viral propagation in the lungs of MDA5−/− mice reveals enhanced replication and a distinct distribution involving the interstitium . Though the levels of antiviral cytokines were comparable early during SeV infection , type I , II , and III IFN mRNA expression profiles were significantly decreased in MDA5−/− mice by day 5 post infection . Taken together , these findings indicate that MDA5 is indispensable for sustained expression of IFN in response to paramyxovirus infection and provide the first evidence of MDA5-dependent containment of in vivo infections caused by ( − ) sense RNA viruses .
Innate pathogen sensors detect viral products and respond by initiating a signaling cascade that leads to rapid anti-viral response involving secretion of type I IFNs ( i . e . IFN-α and IFN-β ) and inflammatory cytokines ( i . e . IL-6 and TNF-α ) [1] . In particular , type I IFNs restrict infection by inhibiting viral replication within cells and by stimulating the innate and adaptive immune responses . Once induced , secreted IFN-α and IFN-β bind to the IFNα receptor on the cell surface in an autocrine or paracrine manner . Activation of this receptor initiates the JAK/STAT signal transduction pathways [2] , [3] and the expression of IFN-inducible genes [4] . These gene products increase the cellular resistance to viral infection and sensitize virally-infected cells to apoptosis [5] . In addition , type I IFNs directly activate DC and NK cells , and promote effector functions of T and B cells , thus providing a link between the innate response to infection and the adaptive immune response [6] , [7] . Several viral sensors have been identified that belong to the Toll-like receptor ( TLR ) and RIG-I like receptor ( RLR ) families [8] . TLRs are expressed on the cell surface and/or in endosomal compartments [9] . TLR3 recognizes double stranded RNA ( dsRNA ) , a molecular pattern associated with replication of single stranded RNA ( ssRNA ) viruses as well as the genomic RNA of dsRNA viruses [10] . TLR7 and TLR8 recognize ssRNA [9] , [11] , [12] , whereas TLR9 recognizes unmethylated CpG-containing DNA [13] . RLRs are cytoplasmic proteins that recognize viral nucleic acids that have gained access to the cytosol [14]–[19] . The RLR family consists of three known members: retinoic acid-inducible gene I ( RIG-I ) , melanoma differentiation-associated gene 5 ( MDA5 ) , and LGP2 . RIG-I and MDA5 both contain a DExD/H box helicase domain that binds dsRNA , a C-terminal domain and two N-terminal caspase recruitment domains ( CARDs ) that are involved in signaling [8] , [17] , [20] , [21] . LGP2 contains a helicase domain but lacks the CARDs , and its precise contribution to antiviral signaling remains ambiguous [17] , [22] . Though RIG-I and MDA5 share common downstream signaling via activation of IPS-1 ( also called MAVS , VISA or Cardif ) and IRF3 [23]–[26] , these helicases exhibit distinct substrate specificity . In this regard , RIG-I has been shown to preferentially recognize ssRNA that is phosphorylated at the 5′ end [27] , [28] and dsRNA molecules which are relatively short [29]–[31] . In contrast , MDA5 recognizes long dsRNAs but does not discern 5′ phosphorylation[30] , [32] , [33] . This distinct ligand preference has been shown to confer specific recognition of individual viruses: RIG-I has been shown to detect Influenza A and B viruses , paramyxovirus , vesicular stomatitis virus ( all ( − ) ssRNA virues ) and some Flaviviruses ( ( + ) ssRNA viruses including Japanese encephalitis virus , Hepatitis C virus and West Nile virus ) [16] , [33] , [34] . In comparison , MDA5 has been shown to selectively detect ( + ) ssRNA viruses including picornaviruses ( encephalomyocarditis virus , Mengo virus and Theilers virus ) [32] , [33] , Caliciviridae ( murine norovirus-1 ) [35] , and Flaviridae ( West Nile Virus and Dengue Virus ) [34] , [36] . Accordingly , it is believed that the presence of different classes of sensors may reflect the need for multiple mechanisms to effectively control the wide variety of viral pathogens . Paramyxoviruses are ( − ) ssRNA viruses that are responsible for a number of human diseases including those caused by measles , mumps , parainfluenza virus and respiratory syncytial virus ( RSV ) . Importantly , infections caused by paramyxoviruses are the most frequent cause of serious respiratory illness in childhood and are associated with an increased risk of asthma [37] , [38] . Sendai virus ( SeV ) is a murine parainfluenza virus which causes an acute respiratory disease in mice that resembles severe paramyxoviral bronchiolitis found in humans following RSV infection [39] . To date , RIG-I is the only dsRNA sensor that has been implicated in the veritable detection of paramyxoviruses [33] , [40] . The importance of RIG-I in the containment of SeV infection is underscored by capacity of SeV C proteins to directly antagonize RIG-I signaling [41] in addition to their ability to inhibit IFN signal transduction [42] , [43] . However , paramyxovirus-encoded V proteins are known to directly interfere with MDA5 function by blocking binding of dsRNA [14] , [44] , thus implicating MDA5 in the containment of paramyxovirus infection as well . In addition , SeV defective interfering ( DI ) particles have been shown to engage MDA5 in vitro [45] , though the in vivo relevancy of this detection mode is unknown . Thus , to determine whether MDA5 functions during natural infection with paramyxovirus in vivo , we assessed mice deficient in MDA5 ( MDA5−/− mice ) following respiratory tract infection with SeV .
In order to assess an in vivo role for MDA5 in containment of paramyxovirus infection , we infected MDA5−/− mice with Sendai virus ( SeV ) . Mice on a C57BL/6 ( B6 ) background were selected for these experiments as the 129 strain is lethally susceptible to SeV at extremely low inocula [46] , thus prohibiting assessment of loss of MDA5 function on this background . A dose of 200 , 000 pfu was administered to mice by intranasal delivery , an infection method that typically results in acute , non-lethal bronchiolitis in B6 mice . As a gross determinant of virus-induced morbidity , % body weight for infected WT and MDA5−/− mice was monitored for 2 weeks post infection ( PI ) . Though essentially identical % weight loss values were observed up until day 8 PI; onwards , weight loss in MDA5−/− mice was significantly more severe ( p<0 . 05 ) ( Figure 1A ) . Correspondingly , histological analysis of lung sections obtained from day 12 PI MDA5−/− mice revealed consolidation of the lung parenchyma as well as notable PAS-positive airway cells , an indication of mucus hyper-secretion ( Figure 1B ) . Severe histopathology was not observed in the lung sections obtained from control mice at this time point . In addition , we compared survival following increasing inocula of SeV ( Figure 1C ) . Though MDA5−/− mice were not susceptible to the 200K pfu SeV dose , MDA5−/− mice fully succumbed to 400K and 600K pfu SeV , between 9–14 days PI . In contrast , control mice were fully resistant to the 400K pfu dose , though 40% mortality was observed for controls infected with the 600K dose . Thus MDA5−/− mice exhibit enhanced morbidity and susceptibility to SeV infection relative to control mice . To more fully assess SeV susceptibility , we extended our analysis of the histological changes seen in the lungs of SeV-infected MDA5−/− mice . H&E stained sections obtained from day 2 PI ( not shown ) and day 5 PI ( Figure 2A ) lungs demonstrated similar patterns of bronchiolitis , though peribronchiolar lymphoid cuffing that formed in the lungs of control mice appeared moderately thicker and more densely populated than those of MDA5−/− mice ( Figure 2A ) . FACS analysis of lung-derived leukocytes at d2 , d5 , and d8 PI revealed no significant differences in lymphoid and myeloid subpopulations ( neutrophils , cDC , macrophage and alveolar macrophage; data not shown ) . Significantly , at d5 and d8 PI , FACS analysis revealed equal relative numbers of lymphoid subpopulations ( CD3+ , CD19+ and NK1 . 1+ ) ; CD69 expression profiles on these subsets were comparable between strains ( data not shown ) . By d8-9 PI , significant pathology was observed in the lungs of SeV-infected MDA5−/− mice ( Figure 2A ) , despite the fact that comparable numbers of SeV-specific CTL were generated in both strains at this time point ( Figure 2B ) . Grossly , lungs dissected from SeV-infected MDA5−/− mice exhibited enhanced areas of hemorrhage relative to control lungs ( data not shown ) . Microscopic analysis revealed epithelial cells that were notably hyperplastic with abundant micropapillary projections . Additionally , severe bronchointerstitial pneumonia was observed , with alveolar walls adjacent to affected airways thickened and congested with chronic inflammatory cell infiltrates and hyperplastic type II pneumocytes , a lung injury pattern consistent with SeV susceptibility [46] , [47] . In comparison , sections obtained from control mice at these later time points exhibited moderate changes to the airway epithelium and mild interstitial infiltration ( Figure 2A ) . As susceptibility to SeV infection correlates with increased viral burden [48] , we next assessed viral replication in wild type and MDA5−/− mice using a combined approach of real-time PCR analysis and specific staining for SeV antigens . Initially , at d2 PI , IF staining of viral antigens in lung sections appeared comparable between the two strains . By d5 PI , SeV antigens exhibited restrained expression in the airways of control mice ( Figure 3A top of panel ) . In contrast , the bronchioles of MDA5−/− mice remained notably positive for SeV antigens at this time point ( Figure 3A bottom of panel ) . More striking however , was the observation that parenchyma tissues proximal to infected airways stained conspicuously for SeV antigens in MDA5−/− mice at d5 PI . In SeV resistant strains of mice , SeV infection is typically restricted to the mucociliary epithelium of the conducting airways , including the trachea , bronchi and bronchioles [49] , [50] . Viral replication that extends to the alveolar spaces is a feature commonly seen in susceptible strains of mice [51] . Accordingly , this pattern of infection supports a role for MDA5 in controlling the replication of SeV during in vivo infection . To confirm this finding , we measured viral RNA levels from WT and MDA5−/− mice infected with 200K pfu SeV using real-time PCR analysis . Assessment was made using primer/probe sets designed to amplify SeV genome ( 3′ untranslated region ) and SeV N gene ( genomic and transcript ) ( Figure 3B and C ) . Using this strategy , an approximate 5 fold increase in SeV genome copy number/Gapdh mRNA was detected on d5 PI , though significant differences were also observed on d2 and d8 PI . Analysis of N gene revealed ∼2 fold increase in expression on days 2 and 5 , though there were no significant differences by d8 PI . Thus it appears that MDA5 contributes in part to the containment of SeV replication in vivo . Though SeV is a potent inducer of type I IFN in the mouse , functioning via several distinct pathways , it possesses several mechanisms by which it can counteract the IFN response . Despite this property , induction of IFN expression [14] , [52] , particularly type I and II , is critical in the containment of SeV infection in vivo as underscored by the profound SeV susceptibility seen for mice deficient in STAT1−/− mice [53] . As MDA5 is known to induce expression of type I IFN in vitro in response to polyI:C stimulation and viral infection [17] , we sought to directly assess the ability of MDA5−/− mice to express IFN in response to SeV infection . In this regard , WT and MDA5−/− mice were infected with 200K pfu SeV and subsequently assessed for cytokine expression by real-time PCR analysis over the acute period . While both strains demonstrated comparable mRNA levels at d2 PI , type I IFN expression was dramatically dampened in MDA5−/− mice at d5 PI ( Figure 4A and B ) . Unexpectedly , significant decreases in expression of Ifn-γ , Il-28b ( Ifn-λ3 ) and Tnf-α mRNA were also observed in the lungs of MDA5−/− mice compared to the WT cohort , with the most dramatic difference observed for Il-28b mRNA expression ( Figure 4C , D and E ) . In contrast , Il-1β , and Il-10 mRNA levels were not significantly different across strains , though the levels of Il-6 mRNA was markedly increased in MDA5−/− mice following infection ( Figure 4F , G and H ) . Accordingly , MDA5 appears to control the expression of SeV-induced anti-viral cytokines , particularly type I , II and III IFNs , during the late acute period ( d5 PI ) , but does not appear to be involved during the immediate early response . Importantly , the decrease in IFN expression coincides with expanded viral propagation in the MDA5−/− mice , suggesting that reduced IFN expression during this time point accounts for the corresponding increased viral burden . Induction of IFN expression transactivates expression of a number of IFN response genes through a signal transduction cascade involving JAK/STAT activation . MDA5 and RIG-I are among the genes induced by IFN signaling in vitro [20] . To determine the expression profile of MDA5 and RIG-I in the airways of mice infected with SeV , mRNA was measured by real-time PCR analysis from whole lung homogenates obtained from WT mice infected with 200K pfu SeV . Expression of Mda5 and Rig-i mRNA was significantly increased at d2 and d5 PI , though the levels began to decline by d8 PI ( Figure 5A ) . Lastly , to determine the tissue distribution of MDA5 expression , lung sections from d5 PI mice were stained with anti-MDA5 polyclonal antibodies . Visualization of MDA5-specific staining was performed using tyramide-based amplification . IF microscopic analysis of affected airways revealed a pattern of MDA5 expression that was primarily restricted to the airway epithelium , though expression was also detected in cells of the proximal interstitum , in particular , in cells that appeared to resemble type II pneumocytes and alveolar macrophage ( Figure 5B ) . Sections from MDA5−/− mice did not stain for MDA5 , confirming the specificity of anti-MDA5 staining . Accordingly these findings indicate that MDA5 is induced following SeV infection and that the lack of expression in MDA5−/− mice accounts for the phenotype described at the later time point .
Our understanding of innate immune factors that recognize and respond to pathogens has greatly expanded over the last decade . A major component of the RNA virus detection system in mammals involves members of the RLR family , including RIG-I , MDA5 , and LGP2 [1] . Elucidating a role for the RLRs in virus-induced IFN production has been facilitated by the availability of RIG-I−/− and MDA5−/− mice [32] , [33] . Initial observations using embryonic fibroblasts and bone marrow derived DCs generated from these mice revealed striking phenotypes including a failure to produce IFN in response to a wide cross-section of viruses and nucleic acids , and an inability to contain viral replication . Specifically , MDA5 was found to be the sole receptor for picornaviruses and caliciviruses ( ( + ) ssRNA viruses ) [32] , [33] , [35] , whereas RIG-I was described as the receptor for ( − ) ssRNA viruses such as paramyxoviruses and orthomyxoviruses , as well as for ( + ) ssRNA viruses belonging to the Flaviridae family [16] , [33] . However , our understanding of these virus recognition systems in vivo is limited , in part because RIG-I−/− mice die perinatally . The precise molecular patterns of virus replication recognized by RIG-I and MDA5 are still not fully clear . Initially , a mimic of viral dsRNA , polyI:C , was found to bind and activate RIG-I . However , ensuing research identified 5′-triphosphate-linked ssRNA as the major RIG-I inducer [27] , [28] . Furthermore , in vitro data obtained using knockout mice suggested in fact that MDA5 , and not RIG-I , recognizes polyI:C , thereby formulating a recognition model whereby RIG-I recognizes short 5′-triphosphorylated RNAs , while MDA5 recognizes dsRNA structures irrespective of the 5′ cap [8] , [30]–[32] . However , more thorough dissection of the helicase binding function and activation process has recently determined that the picture is indeed more complex than previously thought [29] , [40] . In this regard , both helicases have been shown to recognize dsRNA , in a manner that is likely dependent on its length , while RIG-I demonstrates the added ability to respond to 5′-triphosphate ssRNA products . To complicate these paradigms , there is increasing evidence that viruses have evolved various properties aimed at antagonizing or degrading viral sensors . Thus , our understanding of viral recognition by the RLR helicases is evolving . With respect to molecular sensing of paramyxovirus infection , both 5′-triphosphorylated ssRNA and long dsRNA species are likely present in SeV-infected cells , thereby implicating both MDA5 and RIG-I in the antiviral sensing process . However , in vitro studies concur that cultured embryonic fibroblasts and bone marrow-derived DC cells detect SeV RNA chiefly through RIG-I , whereas MDA5 and TLR3 are dispensable [33] , [34] , [41] , [54] . TLR7 and TLR8 in myeloid cells have also been shown to recognize SeV RNA in vitro as well [55] . Regardless , it cannot be excluded that in vivo , other RNA sensors , including MDA5 , may contribute , at least in part , to anti-SeV responses . Indeed , a recent study by Yount et al . has demonstrated that MDA5 can detect SeV DI particles in vitro [45] . The relevancy of this recognition system in vivo is uncertain; certainly in our hands , using SeV/52 , which has a limited ability to form DI particles , as per PCR-based analysis ( data not shown ) , MDA5 appears to exert a significant effect on viral containment . Most importantly however , SeV encodes a V protein that specifically binds to and blocks MDA5 signaling in vitro [14] , [44] . Thus , it is possible that MDA5 does , indeed , detect SeV in vitro , but that it is functionally curtailed by the V protein in this circumstance . Interestingly , in our hands , in vivo infections using SeV with V protein deletion resulted in no real effect on mortality or type I IFN induction across strains ( data not shown ) , likely explained by the fact that deletion of V protein in SeV markedly attenuates virulence and pathogenicty in vivo [56] . While initial characterization of MDA5-deficient cells has not supported a role for MDA5 in containment of SeV , these studies have been limited to observations made in cultured embryonic fibroblasts and in vitro-derived dendritic cells; populations which are not primary targets for SeV replication in the course of the natural infection . Rather , SeV replication mostly occurs in the airway epithelium of the conducting airways [49] , [50] . For these reasons , we hypothesized that SeV propagation would be sensitive to the MDA5 status of the host in the context of an in vivo infection . Indeed , the epithelial cells constitutively express MDA5 at low levels and subsequently up-regulate expression in response to SeV ( Figure 5B ) , a finding that supports the relevance of this RNA helicase to SeV and other airborne infections . Interestingly , MDA5 deficiency did not influence the composition of the inflammatory infiltrate ( data not shown ) , implying that the immune defect is largely restricted to the airway epithelium , the site of viral replication . This is compatible with our earlier findings using STAT-1−/− chimeras , wherein we observed that loss of IFN response in the stromal compartment alone accounted for the immune deficiency to SeV [53] . We therefore sought to further assess the significance of MDA5 in the control of SeV infection in vivo . In this regard we have demonstrated that MDA5 controls SeV replication and spread through induction of type I IFNs , but that this effect appears late ( d5 PI ) , as IFN gene transcription is not impaired on d2 PI ( Figure 4 ) . It is likely that the initial IFN response is sufficient to initiate a range of immune responses , such that the late reduction in IFN transcripts results only in a 2–3 fold change in LD50 ( Figure 1 ) . Whether this specific IFN pattern remains true for other viruses as well remains to be tested . This surprising collapse of the host type I IFN response at d5 PI is accompanied by parallel decreases in the level of Il-28b and Tnf-α expression ( Figure 4 ) , and , curiously , decreased Ifn-γ transcript levels . This later observation may reflect a selective role for MDA5 in the induction of IFN-γ expression by NK cells . Lastly , the MDA5 status does not appear to influence the levels of IL-1β , or IL-10 or the ability of the host to mount a virus-specific CTL response . However , the levels of Il-6 mRNA in whole lung homogenates derived from d2 and d5 PI MDA5−/− was markedly increased , suggesting the induction of compensatory mechanisms in the context of MDA5 deficiency that could potentially account for the enhanced morbidity and mortality seen in the MDA5−/− mice . An additional concern raised by these data is the relative contribution of MDA5 and RIG-I in the response to virus . In light of the existing literature [33] , [40] , it seems likely that RIG-I is responsible for the normal IFN response to SeV early in the infection . Indeed , as depicted in Figure 5A , RIG-I is strongly induced early on during infection . Why the later IFN response depends on MDA5 is not known . MDA5 is encoded by an IFN-upregulated transcript , and it remains possible that it is the accumulation of MDA5 that allows for the subsequent MDA5-dependent IFN response on d5 PI . Yet other IFN-induced genes , notably RIG-I , are also upregulated by IFN , which should provide additional antiviral protection in vivo . Interestingly , SeV encodes a nested set of C proteins that have been shown to impede IFN signaling through direct inhibition of STAT signaling [41] , [42] and which are also known to strongly antagonize RIG-I function [41] . Furthermore , SeV-V proteins have been shown to have direct inhibitory effects on both MDA5 and RIG-I signaling [41] , [44] . Thus it remains possible that the effects of SeV V and C proteins have an accumulative effect on RIG-I function that essentially overwhelms this sensor at d5 PI , and that in this context , MDA5 plays an essential role in containment of SeV . Since assessment of the relative contribution of RIG-I and MDA5 in containment of SeV infection in vivo is not possible , a possible next step in assessing the importance of MDA5 function would involve assessment in MDA5−/− and IPS-1−/− strains . We envision several possibilities that could potentially explain this dramatic effect of MDA5 . The first is that , in the absence of MDA5 , the balance between virus replication and the IFN response is disrupted sufficiently , such that by d5 PI , virus replication has overwhelmed the response in a qualitative fashion –presumably through direct cytotoxic effects or via the overproduction of immunosuppressive C proteins . This possibility is supported by the fact that SeV is replicating to higher levels in the MDA5−/− lung already by d2 PI ( Figure 3B ) . Indeed , in support of this hypothesis , we observe a striking increase in SeV replication that spreads extensively into the interstitium of MDA5−/− lungs compared to controls . Another possible explanation , which we have not assessed , is an apoptotic response potentially mediated by MDA5 . In this scenario , MDA5 would instruct or sensitize infected cells to commit suicide so as to shut down viral replication in infected cells . Indeed , ectopic expression of MDA5 in a melanoma cell line has been shown to inhibit colony formation , presumably through induction of apoptosis [20] , and IPS-1 overexpression induces cell death , as well [57] . In fact , SeV-dependent apoptotic signaling requires IRF3 [58] . In the case of MDA5 deficiency , loss of pro-apoptotic activity could lead to a robust increase in viral replication and enhanced IFN blockade through overexpression of SeV C proteins . This possibility is favored by the fact that , despite a normal IFN response on d2 PI ( Figure 4 ) , the virus is found to be replicating at higher titers ( Figure 3B ) . It seems likely that the inability of the MDA5−/− animals to sustain an IFN response leads to increased viral replication and dissemination on d5 PI , thus causing significantly higher morbidity and mortality in the knockout cohort ( Figure 1 ) . It is important to note , however , that the effects of MDA5 deficiency on SeV replication are much milder than expected if MDA5 were the sole target of the V protein . Indeed , SeV V mutants ( SeV-ΔV ) are severely attenuated; replication is demonstrably abrogated in the lungs by d2 PI [56] . IRF3 deficiency of the host restores SeV-ΔV pathogenicity , suggesting that the mutant virus acts by blocking IRF3 signaling [59] . Yet disease caused by SeV in MDA5−/− mice is milder than the disease seen in the IRF3−/− animals . Consequently , we believe that the V protein must have additional targets besides MDA5 . In this regard , it has recently been demonstrated that Lgp2 encodes a helicase epitope that is akin to the MDA5 helicase , the portion of MDA5 that binds paramyxovirus V proteins [60] , thereby suggesting that LGP2 may be an additional V protein target . In this case , a MDA5-LGP2 double knockout mouse may potentially phenocopy the IRF3 mutation in its response to SeV infection . Taken together , our findings demonstrate that MDA5 significantly contributes to the response to paramyxovirus and constitute the first in vivo demonstration of MDA5 activity against a negative-strand virus . As such , it appears likely that MDA5 has a wider specificity as a viral nucleic acid receptor than initially believed , and that the initial clear-cut cases of either MDA5 or RIG-I being the sole receptor for a given virus will prove to be exceptions rather than rules when studied in the context of in vivo infections .
Control C57BL/6J ( B6 ) mice used in these experiments were purchased from JAX . MDA5−/− mice [32] were backcrossed onto the B6 background to 99 . 9% congenicity . For in vivo SeV infection , Sendai/52 Fushimi strain was instilled intranasally into deeply anesthetized mice and at the indicated time points , mice were humanly sacrificed for harvest of lung tissue . Virus was purchased from the ATCC and subject to two rounds of in vitro plaque purification in Vero cells to eliminate the presence of DI particles . A clone thus identified was then subject to a single round of amplification in embryonated chicken eggs following inoculation of ∼1000 PFU . 24–36 hr post inoculation , SeV was isolated from the allantoic fluids and diluted in phosphate-buffered solution to generate a viral stock that was subsequently characterized on the basis of in vivo infectious properties . Calculation of PFU was performed by standard plaque assay using either Vero E6 cells or LLC-MK2 cells . Importantly , propagation under these conditions does not favor the formation of DI particles , a process that occurs most frequently when virus is repeatedly passaged at high MOI . Indeed , PCR analysis of stock virus indicated the absence of DI genomes . The methods for mice use and care were approved by the Washington University Animal Studies Committee and are in accordance with NIH guidelines . Single cell lung suspensions were made from minced lung tissue subjected to collagenase/hyaluronidase/DNAse I digestion . Staining of surface markers was performed using FcR block and fluorochrome-conjugated mAbs . To immunophenotype the immune infiltrate , specific combinations of mAbs were chosen which discern granulocytes ( Ly6G+ ) , macrophages ( F4/80+ ) , cDC ( CD11c+F4/80−Siglec-H− ) , pDC ( Siglec-H+ CD11cmid ) , NK cells ( NK1 . 1+NKp46+ ) , T cells ( CD3+CD4+/−CD8+/− ) and B cells ( CD19+ ) . SeV-specific PE-labeled pentamer Kb:FAPGNYPAL ( NP 324-332 ) was purchased from Proimmune; cells were stained with CD8 and counterstained with propidium iodide , F4/80 and CD19 to eliminate background . Activation status was determined using specific mAbs for MHC-II , NKG2D and CD69 . Samples were acquired on a FACScalibur ( BD Biosciences ) and analyzed using Cellquest software . RNA was purified from lung homogenate using Trizol Reagent ( Invitrogen ) . RNA was treated with RNAse-free DNAse I ( Ambion ) to eliminate genomic DNA . RNA was converted to cDNA using the High-Capacity cDNA Archive kit ( Applied Biosystems ) . Target mRNA and viral RNAs were quantified by real-time PCR using specific fluorogenic probes and primers and the Fast Universal PCR Master Mix system ( Applied Biosystems ) . Primer sets and probes for mouse Ifn-α2 ( Mm00833961_s1 ) , Ifn-β ( Mm00439552_s1 ) , Ifn-γ ( Mm00801778-m1 ) , Il-28b ( Mm00663660_g1 ) , Tnf-α ( Mm00443259_g1 ) , Mda5 ( Mm00459183_m1 ) , Il-1β ( Mm00434227_g1 ) , Il-6 ( Mm00446190_m1 ) , Il-10 ( Mm00439616_m1 ) mRNA and SeV genome and Gapdh mRNA were purchased from Applied Biosystems . Samples were assayed on the 7500 Fast Real-Time PCR System and analyzed using the 7500 Fast System Software ( Applied Biosystems ) . Levels of specific gene expression were standardized to Gapdh mRNA expression levels . Lungs were perfused and fixed with 4% paraformaldehyde . Tissue was embedded in paraffin , cut into 5 um sections and adhered to charged slides . Sections were deparaffinized in Citrosolv ( Fisherbrand ) , hydrated , and in the case of IF-microscopy , treated to heat-activated antigen unmasking solution ( Vector Laboratories , Inc ) . H&E and PAS sections were visualized by brightfield microscopy . Expression analysis was performed by IF using chicken polyclonal anti-SeV ( Jackson ImmunoResearch Laboratories , Inc ) and rabbit polyclonal anti-mouse MDA5 ( Axxora Life Sciences , Inc ) . Biotinylated secondary antibodies were purchased from Vector Laboratories , Inc ) . SeV and MDA5 expression was visualized using tyramide-based signal amplification with Alexa Fluor 488 or 594 fluorochromes ( Invitrogen ) . Slides were counterstained with DAPI mounting media ( Vector Laboratories , Inc ) . Microscopy was performed using an Olympus BX51 microscope . Real-time PCR data was analyzed using an unpaired Student's t-test . If variances were unequal , Welch's correction was applied . Charted values represent mean ± SEM . Survival statistics were determined using by Kaplan-Meier analysis of paired cohorts . P values below 0 . 05 were regarded as being significant for all analyses . Experiments were repeated a minimum of three times . | The innate immune system possesses an array of sensory molecules which are purposed in detecting viral nucleic acids . Our understanding of how these molecular sensors detect viral nucleic acids continues to evolve . Herein , we demonstrate that MDA5 , a member of the RIG-I-like receptor family , is involved in the detection of paramyxovirus infection in vivo . Specifically , MDA5 appears to trigger antiviral cytokines that inhibit paramyxovirus replication . In this regard , mice that are deficient in MDA5 are unable to express sustained levels of these cytokines and thus succumb to extensive viral propagation and disease . Our findings are largely discordant from previous in vitro studies using cultured cells , where it has been shown that RIG-I and not MDA5 is involved in the innate response to negative sense RNA viruses . Thus , our data provides strong evidence of MDA5-based detection of negative sense RNA viruses , and furthermore underscore the importance of organism-based analysis of the innate system . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"virology/animal",
"models",
"of",
"infection",
"virology",
"immunology/innate",
"immunity",
"infectious",
"diseases/viral",
"infections",
"infectious",
"diseases/respiratory",
"infections",
"immunology/immunity",
"to",
"infections",
"virology/host",
"antiviral",
"responses"
]
| 2010 | Melanoma Differentiation-Associated Gene 5 (MDA5) Is Involved in the Innate Immune Response to Paramyxoviridae Infection In Vivo |
Surface proteins of the obligate intracellular bacterium Rickettsia typhi , the agent of murine or endemic typhus fever , comprise an important interface for host-pathogen interactions including adherence , invasion and survival in the host cytoplasm . In this report , we present analyses of the surface exposed proteins of R . typhi based on a suite of predictive algorithms complemented by experimental surface-labeling with thiol-cleavable sulfo-NHS-SS-biotin and identification of labeled peptides by LC MS/MS . Further , we focus on proteins belonging to the surface cell antigen ( Sca ) autotransporter ( AT ) family which are known to be involved in rickettsial infection of mammalian cells . Each species of Rickettsia has a different complement of sca genes in various states; R . typhi , has genes sca1 thru sca5 . In silico analyses indicate divergence of the Sca paralogs across the four Rickettsia groups and concur with previous evidence of positive selection . Transcripts for each sca were detected during infection of L929 cells and four of the five Sca proteins were detected in the surface proteome analysis . We observed that each R . typhi Sca protein is expressed during in vitro infections and selected Sca proteins were expressed during in vivo infections . Using biotin-affinity pull down assays , negative staining electron microscopy , and flow cytometry , we demonstrate that the Sca proteins in R . typhi are localized to the surface of the bacteria . All Scas were detected during infection of L929 cells by immunogold electron microscopy . Immunofluorescence assays demonstrate that Scas 1–3 and 5 are expressed in the spleens of infected Sprague-Dawley rats and Scas 3 , 4 and 5 are expressed in cat fleas ( Ctenocephalides felis ) . Sca proteins may be crucial in the recognition and invasion of different host cell types . In short , continuous expression of all Scas may ensure that rickettsiae are primed i ) to infect mammalian cells should the flea bite a host , ii ) to remain infectious when extracellular and iii ) to infect the flea midgut when ingested with a blood meal . Each Sca protein may be important for survival of R . typhi and the lack of host restricted expression may indicate a strategy of preparedness for infection of a new host .
Rickettsia ( Rickettsiales: Rickettsiaceae ) are Gram-negative , obligate intracellular bacteria that are maintained in enzootic cycles involving both hematophagous arthropod vectors and vertebrate hosts [1] . Rickettsiae are the causative agents of significant human diseases such as Rocky Mountain spotted fever ( R . rickettsii ) and epidemic typhus ( R . prowazekii ) . Classically , the members of the genus Rickettsia have been divided into two groups: the tick-transmitted spotted fever group ( SFG ) and the insect-transmitted typhus group ( TG ) based on their antigenic and molecular profiles . However , these groups share some antigenic proteins such as outer membrane protein B ( OmpB ) and 17 kDa lipoprotein [2] . The tick-transmitted SFG currently includes over 16 species , several of which are known human pathogens ( R . rickettsii , R . conorii , and R . sibirica ) . The louse and flea transmitted TG rickettsia contain the pathogenic species R . prowazekii and R . typhi ( causative agent of murine typhus ) . Extensive phylogenetic and comparative genomic analyses have resulted in the proposal of the ancestral group ( AG ) and transitional group ( TRG ) rickettsia and these include species with mild or unknown pathogenicity as well as broad arthropod host ranges [3] , [4] , [5] . Despite the recent advances made in rickettsial molecular biology and genomics , their determinants of pathogenicity still remain undefined . Because of the involvement of rickettsial Omps in cell surface recognition , initial binding of bacteria to host cells , invasion processes [6] , [7] as well as their immunogenicity and utility as vaccine candidates , this group of proteins has been a target of interest [8] , [9] , [10] , [11] . Bacterial surface-exposed proteins are involved in an array of processes including sensing the environment , protection from environmental stresses , adherence to and invasion of host cells , cell growth and interaction with the immune system . For intracellular bacteria , surface exposed proteins interact with host cytoplasmic or organelle proteins [12] , [13] . Characterizing the surface composition of rickettsiae allows for identification of factors required for successful colonization of mammalian and arthropod hosts . The family of rickettsial autotransporters ( ATs ) are referred to as surface cell antigen ( Sca ) proteins [14] . A previous computation analysis identified 17 orthologs Sca0 ( OmpA ) , Sca1-Sca4 , Sca5 ( OmpB ) , Sca6-Sca16 ) distributed throughout nine complete rickettsial genome sequences [15] . These genomes as well as those subsequently sequenced , each encode a unique repertoire of sca orthologs that are present in different functional states ( i . e . complete , fragmented or split ) . Furthermore , while the AT domains are well conserved within orthologs and to a lesser degree across paralogs , much less amino acid identity is observed among the passenger domains [15] . Experimental studies have detected expression of Sca0 , Sca5 and Sca4 by various methods [16] , [17] , [18] , however , only Sca0 , Sca5 and more recently Sca1 and Sca2 of R . conorii have been shown to function as adhesins [6] , [19] , [20] , [21] . The internal repeat motifs within the passenger domains are proposed to render each protein an adhesin and contribute specificity to host receptors [15] . For R . conorii Sca5 , these repeats may contribute to its specificity for the DNA protein kinase Ku70 on the surface of host cells [22] . More recently , Sca2 in R . parkeri has been characterized as a formin-like mediator of actin-based motility [23] indicating that some of the Scas have intracellular functions and may interact with host proteins to promote rickettsial survival . R . typhi , the focal organism of this study , contains 5 sca orthologs in its genome – sca1 , sca2 , sca3 , sca4 and sca5 . As in other species studied , Sca5 likely mediates adherence and invasion of R . typhi to host cells , however little is known about the expression and function of the other Sca orthologs during R . typhi infection in either mammalian or arthropod hosts . A better understanding of the expression and distribution of the Scas during R . typhi transmission and infection is crucial in order to appreciate the function of these proteins . This study is a comparative analysis that was undertaken to elucidate the transcriptional and protein expression profiles of the R . typhi Sca family in vitro ( tissue culture ) and in vivo ( rat and flea infections ) . This study addresses our hypothesis that R . typhi Sca expression is time and host dependent .
Surface proteins of intracellular bacteria mediate interactions required for their pathogenesis and survival . Defining the surface proteome of R . typhi provides a better understanding of the potential interactions at the host-pathogen interface . Using CoBaltDB [24] and pSORTb v . 3 . 0 [25] , several signal peptide and subcellular localization algorithms were employed to generate predictions of secreted and outer membrane proteins among the 838 ORFs in R . typhi . Of the 838 ORFs , 140 were predicted to be secreted by at least one of the four algorithms employed ( Figure 1 ) . Further , 25 proteins had no detectable secretion signals but were predicted to be localized to the membrane or extracellular to bacteria ( Table S1 , proteins highlighted in yellow ) . Only the SOSUI-GramN [26] and PSORTb v . 3 . 0 [27] algorithms identified proteins localized to the cytoplasmic membrane , inner membrane or outer membrane ( Tables S1 and S2 ) . In a previous study , putative secretion signals R . typhi ORFs were tested for the ability to facilitate secretion to the periplasm via a sec-translocon dependent mechanism in E . coli [28] . We further analyzed all R . typhi str . Wilmington ORFs using the SecretomeP method [29] which identifies non-classically secreted proteins using a sequence-derived feature based approach and was trained on proteins experimentally identified on the bacterial surface but not predicted to be secreted by SignalP algorithms . Based on the SecretomeP method , 26 of the proteins with classical secretion signals also have signals for non-classical secretion methods ( Figure 1 and Table S1 , proteins highlighted in green ) . Fifty-four proteins were predicted to have only non-classical signals ( Figure 1 and Table S2 ) . Only a fraction of the proteins predicted to be secreted were identified in the surface proteome analysis . A total of 68 proteins were detected ( Table S3 ) , 27 of which are predicted to be secreted ( to the periplasm via the Sec translocon as determined by SignalP ) or surface-exposed ( in the outer membrane or extracellular to bacteria as determined by SOSUIGramN and pSORTb ) by at least one of the algorithms used . Fifteen of the proteins are predicted to have secretion signal peptides as determined by SignalP NN and/or HMM algorithms . pSORTb v . 3 . 0 [25] predicted three proteins to be localized to the outer membrane [RT0565 and RT0699 ( Sca5 ) ] or be extracellular ( RT0522 ) . SOSUIGramN localized 11 proteins to either the outer membrane ( RT0521 , RT0744 and RT0805 ) or extracellular [RT0052 ( Sca2 ) , RT0699 ( Sca5 ) , etc . ] to bacteria ( Table S3 cells shaded grey ) . LipoP , Phobius and SecretomeP were the only algorithms to predict signal sequences for one ( RT0117 ) , two ( RT0222 and RT0584 ) and five ( RT0138 , RT0176 , RT0362 , RT0485 and RT0638 ) proteins , respectively ( Tables S3 ) . The predicted signal peptides for 11 of these 68 proteins were previously tested for Sec-translocon dependent secretion into the periplasm using an alkaline phosphatase ( PhoA ) gene fusion system ( see footnotes Table S3 ) and 10 of them were found to mediate secretion [28] . Proteins predicted by the SecretomeP method may be present in the outer membrane as a result of Sec-independent secretion mechanisms . Most of the peptides identified are predicted to be cytoplasmic or have unknown localizations with few of them predicted to be localized to the surface of the bacterium as determined by pSORTb and SOSUIGramN . The expression of cytoplasmic proteins on the surface of bacteria is not uncommon . Of the proteins identified in this study , 18 have homologs that have been shown to be surface exposed in other rickettsial species [30] , [31] , [32] , [33] , other Gram-negative bacteria [34] , [35] , [36] and Gram-positive bacteria [37] . Additionally , 14 of the proteins are annotated as hypothetical and only 5 of the 14 have predicted signal peptides detected by at least one of the algorithms further suggesting that rickettsial proteins may have novel secretion signals . The identification of most of the Sca proteins on the surface was largely consistent with the localization predictions ( Table S1 ) . Specifically , both pSORTb algorithms , LipoP , Phobius and SecretomeP predicted the presence of signal peptides for Sca1-3 and Sca5 . SOSUI-GramN and pSORTb also localized each of these proteins as extracellular to the bacteria or to the outer membrane . SecretomeP was the only program to predict a signal sequence within Sca4 ( Table S2 ) ; it is also predicted to be localized to the cytoplasm by the SOSUI-GramN and pSORTb algorithms . Sequence analysis of Sca1-Sca5 of R . typhi indicates that all five proteins are full length and contain characteristics typical of other Scas from rickettsiae ( Figure 2A ) . Repeat regions were predicted within only three of the five proteins ( Sca2 , Sca3 and Sca4 ) . Consistent with a sequence analysis of Sca2 from R . parkeri [23] , Sca2 of R . typhi contains a proline-rich tract and a series of five WH2 domains; however , the position of these motifs within the passenger domain of R . typhi ( as well as R . bellii , R . prowazekii and one of the two R . akari proteins ) differs greatly from R . parkeri and the other orthologs of SFG Rickettsiae ( Text S2 ) . In general , the passenger domains from Sca1 , Sca2 , Sca3 , and the full Sca4 protein differ in sequence length , percent amino acid identity , and number of repeat regions across the 16 analyzed Rickettsia taxa ( Figure 2B ) . While the distribution of the Sca paralogs across these 16 taxa somewhat correlates with the four rickettsia groups , only the R . felis genome encodes a full-length ortholog for each of the five R . typhi Scas ( Text S2 and Figure S1 ) , which like R . typhi , is vectored principally by fleas . We investigated possible differential transcription of sca genes during L929 cell infection . Of the five sca genes annotated in the R . typhi str . Wilmington genome , only sca4 and sca5 were known to be expressed at the transcript level before this study . All five sca gene transcripts were detected in total RNA extracted from L929 fibroblasts infected for 0–120 hours ( Figure 3A ) . This complements data from a previous study in which sca2 , sca4 and sca5 were identified in a genome-wide screen for temperature-shifted genes in vitro [38] . In general , a decrease in expression was observed within 1 h of infection and median expression levels near to or above initial infection were observed by 120 h ( Figure 3A ) . Rickettsiae were actively growing over the course of the experiment ( Figure 3B ) . Initial analysis using the OperonDB algorithm predicted that sca3 , sca4 and sca5 genes had a probability of co-occurring in the same operon with a confidence of at least 73 ( Table S4 ) ; however , analysis using the updated algorithm resulted in greatly reduced probabilities that any of these sca genes was located within an operon . However , sca3 , sca4 , and sca5 are each positioned downstream of other open reading frames ( Figure 4A ) . Therefore , we hypothesized that each gene cluster was an operon and that each sca was co-transcribed with the upstream genes . We tested the above hypothesis , despite the low prediction probabilities , using RT-PCR to amplify the indicated regions ( Figure 4A ) . Co-transcription of Sca3 with the gene encoding the protease Lon and Sca4 with one of four ATP/ADP translocases , tlcD , were detected ( Figure 4B ) indicating that both of these genes comprise a transcriptional unit with their nearby genes respectively . Co-transcription data for the putative Sca5 operon is less conclusive . We were unable to detect a full transcript spanning the ihfA to Sca5 genes ( data not shown ) and therefore designed primers to amplify 5 different regions that would span the entire putative operon . Transcripts of Sca5-spoTd and spoTd-RT0701 , are detected suggesting that these 3 genes are co-transcribed . While an amplicon for RT0701-RT0702 is detected , a transcript spanning RT0702 and spoTd could not be amplified ( Figure 4C ) . Multiple primer combinations were tested for the former combination without success . To investigate Sca protein expression we generated a rabbit polyclonal antibody ( pAb ) against a peptide within the passenger domain of each Sca . Specificity of the serum directed to each Sca was confirmed by immunoblot against lysates from uninfected and infected L929 cells ( Figure S2 ) and sections of infected and uninfected L929 cells ( Figure S3 and S4 ) . Sca5 staining is observed in the cytoplasm and the outer membrane on many rickettsiae ( Figure S3 ) . Some labeling of host cell chromatin was also observed . Although staining is weak , Sca 1 is observed in the rickettsial cytoplasm and in association with the membrane of rickettsiae . Staining is also observed on the outer membrane ( Figure S3 ) , the host cell cytoplasm and chromatin ( data not shown ) . Sca2 and Sca3 staining is observed on the outer membrane , host chromatin or cytoplasm of very few rickettsiae ( Figure S3 ) . Sca4 staining shows a similar pattern to the other Scas with labeling of the outer membrane , periphery and rickettsial cytoplasm ( Figure S3 ) . To investigate possible variability of Sca expression within organisms that maintain the infection in an enzootic cycle we infected rats and cat fleas . Rats serve as a primary reservoir for murine typhus in urban environments . Infection is self-limiting , however , splenic dissemination occurs a week into infection [39] . The cat flea , C . felis , is suspected of being one of the primary vectors of R . typhi transmission to small mammals and humans in the United States [40] . No staining was observed in sections from uninfected fleas or uninfected rat spleens ( Figures S5 and S6 ) . Scas3-5 are expressed in rickettsiae in various organs of 14-day infected fleas including the midgut and developing eggs ( Figure 3C ) ; staining for Scas 1 and 2 were not conclusive . Positive staining was distinctly detected for Scas1 and 5 by immunofluorescence staining of cryosections of spleens from 9-day infected Sprague Dawley rats ( Figure 3D ) indicating that they are expressed during mammalian infection . However , staining for Sca4 was diffuse and not clearly detected ( data not shown ) . Expression of the Scas on whole rickettsiae was investigated using flow cytometry and negative stain IEM of whole rickettsiae . For flow cytometry , rickettsiae were stained with antibodies prior to fixation to avoid permeabilization so that only surface-bound Scas were detected . While the majority of rickettsiae only stained positive for the anti-R . typhi sera , a small proportion ( 1 . 58–3 . 82% ) also stained positive for the respective Sca protein ( Figure S7A ) . Non-specific staining of rickettsia and any co-purified host components was minimal ( Figure S7B ) . To confirm surface expression of the Sca proteins , whole rickettsiae placed onto grids were labeled with anti-Sca sera followed by colloidal gold conjugated anti-rabbit IgG secondary antibody . Immunogold labeling for each Sca protein is observed on the surface of negatively stained , intact rickettsia with clusters of gold particles indicating positive staining ( Figure 5 ) .
Surface proteins of obligate intracellular bacteria comprise a crucial interface for pathogen-host interactions by mediating the initial attachment and infection of host cells and subsequent contact with host cytosolic proteins to promote bacterial survival and replication through the subversion of host processes . In this report , we investigated the surface proteome of R . typhi str . Wilmington , first by bioinformatic analyses to predict subcellular localizations and the presence of classical and non-classical secretion signals , then by selective labeling and purification of surface proteins for identification . We consequently focused on characterizing the Sca family of autotransporters . Outer membrane proteins ( OMPs ) are immunodominant in rickettsial infections and immunization with these antigens has been shown to confer protection from lethal challenge in animal models [8] , [9] , [10] , [11] , [41] . Similarly , the Major Surface Proteins ( MSPs ) of Anaplasma spp . and the Outer Membrane Proteins ( OMPs ) of Ehrlichia spp . are also identified as immunogenic rickettsial proteins [31] , [32] , [42] . Rickettsial OMPs form the basis for antigenic relationships between and within phylogenetic groups and allow for some cross-protection from infection by multiple species . A major basis for the antigenic similarity of Rickettsia spp . is the presence of the complete gene for the 120 kDa outer membrane protein B ( Sca5 ) in all species [15] . Bioinformatic analyses place Sca5 and the similar protein OmpA , which is not encoded in the genomes of TG Rickettsia , into the superfamily of proteins known as the type V secretion system ( TVSS ) or autotransporter family [43] . Many of the proteins identified on the surface of R . typhi ( Table S3 ) have homologs that were similarly identified on the surface of other rickettsiae or other families of Gram-negative bacteria . These include the chaperone proteins GroEL and DnaK and enzymes pyruvate decarboxylase and fumarase ( Table S3 ) . Furthermore , proteins experimentally determined to be surface localized were identified in this study but many currently have no assigned function ( i . e . hypothetical proteins ) in the context of rickettsial infection . Homologs of proteins predicted to be cytoplasmic but experimentally localized to the surface and cytoplasm of other Gram-negative and Gram-positive bacteria [34] , [36] , [44] are detected on the surface of rickettsiae in this study [30] , [31] , [33] . Little is known about the significance of such proteins on the bacterial surface or the mechanism ( s ) by which they are targeted to the surface . However , it is becoming understood that such proteins have alternative functions when surface-exposed and have thus been termed moonlighting proteins [45] For example , when at the bacterial surface , the chaperone protein DnaK of Mycobacterium tuberculosis has been shown to bind plasminogen , stimulate chemokine synthesis in dendritic cells and compete with the human immunodeficiency virus ( HIV ) coreceptor chemokine receptor 5 ( CCR5 ) [46] , [47] , [48] . Elongation factor-Tu ( EF-Tu ) and the E1 beta subunit of the pyruvate dehydrogenase ( Pdh ) of M . pneumoniae has been shown to bind fibronectin [49] and a fungal dihydrolipoamide dehydrogenase , the E3 subunit of Pdh , has been characterized as an acetyltransferase both when not in the cytoplasmic environment [50] . Few rickettsial surface-exposed proteins have been investigated; however , certain Sca proteins have been localized to the bacterial surface [23] or inferred to be located there based on antibody inhibition of infection or conference of adhesive and invasive properties to recombinant E . coli [19] , [20] , [51] . We were able to detect the recently characterized RT0522 conserved hypothetical protein , which encodes a phospholipase A2 homolog that was found to be secreted from the bacterium into the host cytosol [52] . Most of the algorithms used in this study detected no signal peptide in RT0522 and predicted it to be localized to the cytoplasm; however pSORTb v3 . 0 predicted it to be extracellular . This may be evidence that other hypothetical proteins identified on the surface but predicted by consensus to be cytoplasmic and lacking detectable signal peptides , may in fact have extracellular functions as effector proteins . It is also an indication that the algorithms , such as SubcellPredict , SLPLocal and SubLoc v . 1 [53] , [54] , [55] ( data not shown ) , which are trained on free-living model Gram-negative and Gram-positive bacteria , may not accurately identify the signals and motifs utilized by rickettsiae to target proteins to the surface . Discrepancies between the number of predicted outer membrane or extracellular proteins and the number actually identified may be due to low abundance of some peptides and or inaccessibility to the biotinylation reagent . It is predicted that Sca5 comprises the S-layer of rickettsiae and constitutes 15% of the total protein mass [56] , [57] and might block many potential interactions . The properties of the labeling reagent must also be taken into account . The N-hydroxysulfosuccinimide ( NHS ) group of Sulfo-NHS-SS-biotin specifically reacts with the ε-amine of lysine residues and the reactions also proceed at a wide range of temperatures . A more comprehensive analysis of the surface proteome might be compiled by performing reactions across the range of temperatures and using similar biotinylated reagents with spacer arms of differing lengths and reactivity with different groups . The Sca family of autotransporters drew our focus because their primary involvement in adherence and invasion is becoming more apparent but predicted functional domains within a few of them point to other roles [20] , [51] , [58] . Multi-functional autotransporters are not uncommon and further characterization of this family may define distinct extracellular and intracellular roles . For instance , R . parkeri Sca2 is observed to act as an actin-assembly mediator that mimics the eukaryotic formin proteins . Furthermore , it is the first bacterial protein to be identified to functionally and perhaps structurally mimic a domain that was previously thought to be solely eukaryotic [23] . Interestingly , this is the second instance of a rickettsial protein containing a domain with a nearly strict eukaryotic distribution . Similarly , the Sec7-domain-containing proteins ( RalF ) encoded within Rickettsia and Legionella species are unknown from other prokaryotes [59] . As stated above , all of the Scas in R . typhi were identified on the surface when considering the biotin labeling ( Table S3 ) , negative staining electron microscopy ( Figure 5 ) and flow cytometry data collectively ( Figure S7 ) . All Sca proteins except Sca4 were predicted to have signal peptides and be localized to the outer membrane or be extracellular ( Tables S1 and S2 ) . The prediction of a signal peptide in Sca4 by SecretomeP supports evidence that the first 200 bp of Sca4 encode sufficient information to direct secretion of a phoA fusion peptide across the inner membrane into the periplasm ( unpublished data ) . Further , Sca4 is known to generate an antibody response to infection [60] and others have shown , as we have here , that Sca4 is under positive selection [15] . It has been posited that Sca4 and the similarly AT domain-less Sca9 may be exported through the AT domain of another Sca . However a quick analysis for possible signals based on homology among the Scas returned no supporting data for this hypothesis . Further investigation of the function and secretion mechanism of Sca4 is necessary . Phylogeny estimation of Sca1 , Sca2 , Sca4 , and Sca5 proteins resulted in trees that all differ from the Rickettsia species tree ( Figure S1 ) . This is not unexpected , as phylogenies based on single rickettsial genes and proteins rarely agree with those estimated from multiple molecules [61] . Notwithstanding , the Scas are large proteins and contain many variable sites , which should provide enough information for robust phylogeny estimation . However , the extracellular location of the passenger domains likely orchestrates selective pressures on these proteins that render them evolutionarily divergent from the species historical trajectory . Indeed , previous studies have identified positively selected sites within the Sca passenger domains [15] , [62] . For Sca1 , Sca2 and Sca5 , we estimated separate trees based on the passenger domain and AT domain to determine conflicting phylogenetic signals within these domains ( Figure 6A–C ) . In each case , the trees estimated for the AT domain were much more consistent with the rickettsial species tree ( Figure 1B ) , while the trees generated from the passenger domains clearly illustrate Sca protein diversification inconsistent with rickettsial species phylogeny . Recombination between Sca orthologs from different species has also been demonstrated to play a role in the diversification of these proteins [62] . While recombination has been observed experimentally in Rickettsia [63] , [64] , its contribution to rickettsial diversity is currently not fully understood . Notwithstanding , all rickettsia genomes encode enzymes involved in homologous recombination [65] , so it is likely that active recombination occurs across species and strains . Insight into how recombination may shape Sca protein evolution is provided by a phylogeny estimation of an expanded set of Sca4 proteins ( Figure 6D ) . Without an AT domain , contrasting phylogenetic signals within Sca4 proteins could not be determined as for the other Scas described above . However , the identification of an ancestral lineage of plasmid encoded Sca4 ( Figure 6D , yellow box ) proteins illustrates the inclusion of these proteins in the rickettsial mobilome ( all the mobile genetic elements in the genome ) , allowing for the dissemination of Sca variants via lateral gene transfer ( LGT ) . Other studies have reported the presence of Sca ORFs and fragments present on diverse rickettsial plasmids [66] , [67] further supporting the role of LGT as a facilitator of mosaicism via recombination across divergent Sca orthologs ( and possibly paralogs ) . Like the phylogenies estimated from the other Scas , the Sca4 tree does not corroborate the rickettsial species tree , and it is likely that all Scas are subject to positive selection and recombination . These selective forces , which are counter to the evolutionary history of rickettsia , are problematic for inferring species relatedness [62] . Thus , despite their usefulness as rickettsia-specific diagnostic markers , as well as their increasing accumulation from published studies , the Scas should be avoided for phylogenetic inference and classificatory purposes . Transcriptional analysis in non-phagocytic cells demonstrates that sca transcription is sustained during infection ( Figure 3A ) . Although we are unable to correlate sca transcription with protein levels at present , the constitutive expression of all sca genes points toward essential functions . Moreover , the co-transcriptional analyses ( Figure 4 ) suggest that scas and the genes that are co-transcribed may function in the same processes or are part of regulatory mechanisms important for their expression . The inability to detect ihfA-RT0702 and RT0702-spoTd transcripts while detecting RT0702-RT0701 and RT0701-spoTd transcripts in the proposed sca5-ihfA operon indicates that the consecutive genes are not expressed at the same level . It is becoming understood that the concept of multiple successive genes under the control of a single promoter - that is operons - does not inherently mean that all genes are equally expressed [68] . This phenomenon , termed operon polarity , may be explained by the presence of internal transcription terminators , the activity of small RNAs ( sRNAs ) or by riboswitches within operons that modulate transcription in response to metabolite binding [69] . We attempted to detect all of the Sca proteins for R . typhi in the spleen tissue , however , in these experiments we consistently observe expression for Scas 1 and 5 . This may be specific to the rat we used in this particular experiment or the time point we chose to assay ( 9 days ) . Given animal to animal variation as well as time-dependent expression of many rickettsial proteins , we do not exclude the possibility that the remaining Scas play an important role during mammalian infection . While we show that the proteins are expressed in vivo ( Figure 3B–C ) , we are unable to quantitatively assess this expression and correlate it with different stages of growth or infection . However , the immunogold labeling of host cell cytoplasm and chromatin may suggest translocation of the rickettsial proteins to the cytosol . Probing infected cells with pre-immune sera or uninfected cells and tissues with Sca immune sera show no cross-reactivity with host cell components . Proteins may be expressed but not required for infection of a particular host or cell type . Rather , the bacteria will be prepared , upon exiting a cell , to infect an endothelial cell or the flea midgut epithelium . Further characterization of these proteins and identification of interacting host proteins will determine where and how these proteins function and their importance in infection of mammalian and arthropod hosts .
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 Institutional Animal Care and Use Committee ( IACUC protocol no . 1108009 ) of the University of Maryland , Baltimore ( assurance number A3200-01 ) . Rickettsiae were grown and maintained as previously described [38] . Low passage mouse fibroblast cells ( L929 , ATCC CCL1 , ATCC , Manassas , VA ) were grown in Dulbecco's Modified Eagle's Medium ( DMEM ) supplemented with 5% FBS at 37°C and 5% CO2 in 150 cm2 vented lid flasks . When the cells were 80% confluent they were infected with R . typhi str . Wilmington at a multiplicity of infection ( MOI ) of 10 . To harvest rickettsiae , infected L929 cells were scraped into the media then sonicated at setting 6 . 5 twice for 30 seconds using a Sonic Dismembranator ( Fisher Scientific , Pittsburgh , PA ) . The lysates were centrifuged at 1000× g for 5 min to remove large host cell material . The supernatant was centrifuged at 14 , 000× g for 10 min and the pellets , containing rickettsiae , resuspended in 1 mL SPG buffer ( 218 mM sucrose , 3 . 76 mM KH2PO4 , 7 . 1 mM K2HPO4 , 4 . 9 mM potassium glutamate ) . The rickettsiae were placed over a 20% OptiPrep Density Gradient medium:SPG bed ( Sigma-Aldrich , St . Louis , MO ) and centrifuged at 14 , 000× g for 10 min . The pellets were washed twice in SPG buffer and centrifuged at 14 , 000× g; rickettsiae were quantified using the BacLight Live/Dead assay ( Molecular Probes , Eugene , OR ) as per the manufacturer's instructions and stored at −80°C . For flow cytometry , infected host cells were scraped into the media , centrifuged at 14 , 000× g for 10 min then resuspended in 5 ml SPG buffer . Suspended host cells were ruptured by passage through a 27-gauge needle three times to disrupt host cell membranes while maintaining rickettsial membrane integrity and lysates were centrifuged at 1000× g for 5 min . Supernatants were placed over a 20% OptiPrep:SPG bed and centrifuged as above . Bacterial pellets were resuspended in 100 µl SPG . Sub-confluent monolayers of L929 fibroblasts were grown in 150 cm2 vented-lid flasks and infected with renografin-purified rickettsiae at an MOI of 10 for 48 h when most cells are heavily infected . Two flasks were used for each treatment . Rickettsiae were partially purified as described above , washed three times in ice-cold PBS ( pH 8 . 0 ) and pelleted using centrifugation at 8 , 000× g for 3 min at 4°C . Pellets were resuspended in 960 µl of PBS with 80 µl of 10 mM EZ-Link Sulfo-NHS-SS-Biotin ( Pierce Thermo Scientific , Rockford , IL ) and incubated on ice for 30 min . As a negative control , rickettsiae were resuspended in PBS only to assess background affinity purification . As a control for background host material , a mock partial purification was performed on uninfected L929s , and the resulting pellets were labeled as described above . Free biotin was quenched by washing rickettsiae once in 50 mM Tris-HCl ( pH 7 . 5–8 . 0 ) followed by two washes in ice-cold PBS and resuspended in RIPA buffer ( 25 mM Tris-HCl [pH 7 . 6] , 150 mM NaCl , 1% NP-40 , 1% sodium deoxycholate , 0 . 1% sodium dodecyl sulfate ) supplemented with 1X Halt Protease Inhibitors ( Pierce Thermo Scientific ) . Rickettsiae were lysed by bead beating with ≤106 µm glass beads ( Sigma-Aldrich , St . Louis , MO ) . Beads were removed by centrifugation at 500× g for 3 min and cell debris pelleted by centrifugation at 16 , 000× g for 10 min . Supernatants were stored at −80°C until needed . Labeled proteins were purified by affinity purification over columns containing NeutrAvidin Agarose resin ( Pierce Thermo Scientific ) as previously described [42] with modifications . Briefly , columns were equilibrated with three column volumes of wash buffer A ( 25 mM Tris-HCl pH 7 . 6 , 0 . 15 M NaCl , 0 . 5% NP-40 , 0 . 5% sodium deoxycholate , 0 . 05% SDS ) . Biotinylated samples were allowed to enter the resin bed and incubated at room temperature for 10 min . Unbound proteins were washed away with two column volumes of buffer B-1 ( 25 mM Tris-HCl pH 7 . 6 , 0 . 65 M NaCl , 0 . 1% NP-40 ) , then one volume of buffer B-2 ( 25 mM Tris-HCl pH 7 . 6 , 1 . 15 M NaCl , 0 . 1% NP-40 ) followed by one volume of Tris-HCl buffer ( 25 mM Tris-HCl , 0 . 15 M NaCl ) . Washes were collected in 0 . 5 ml fractions and stored on ice . Captured proteins were eluted with two volumes of 5% β-mercaptoethanol-PBS . The eluates were pooled and total protein determined using a BCA assay ( Pierce Thermo Scientific ) . Eluates were concentrated by filtration using Amicon-Ultra prior to addition of 2X SDS sample buffer ( Invitrogen , Carlsbad , CA ) then stored at −80°C . NeutraAvidin ( Pierce Thermo Scientific ) affinity purified proteins were separated on 4–20% Tris-glycine gels ( Invitrogen ) and stained for total protein visualization using PS-Blue ( BridgePath Scientific , Frederick , MD ) . Proteins from several regions were identified using LC-MS/MS at the University of Maryland Baltimore Proteomics Core Facility as described fully in the supplementary methods . Briefly , Coomassie-stained protein bands were excised , dehydrated , digested and de-salted in preparation for LC MS/MS . MS/MS spectra were searched against a uniprot mouse database ( uniprot release number 2010_05; 64 , 389 sequences ) and a Rickettsia typhi database ( uniprot release number 2010_05; 1 , 676 sequences ) using Sorcerer-SEQUEST ( SageN Research , Milpitas , CA ) . The quality of peptide and protein assignments was assessed using PeptideProphet and ProteinProphet . Proteins with probabilities ≥0 . 9 were accepted as true positive identifications . Proteins identified by one unique peptide were manually verified . Identified proteins were then analyzed for signal sequences and motifs that predict subcellular localization . Signal peptides were detected using the SignalP 3 . 0 and LipoP 1 . 0 servers which predict the presence and location of signal peptide cleavage sites in peptide sequences [70] , [71] . SignalP predictions are based on a combination of artificial neural networks ( NN ) and hidden Markov models ( HMM ) while LipoP predictions distinguish between lipoprotein signal peptides , other signal peptides and N-terminal membrane helices . The Phobius server [72] was also utilized primarily for the prediction of signal peptides in an amino acid sequence . To predict subcellular localization predictions for peptides identified on the surface , we employed the pSORTb v3 . 0 . 2 server and the predictive program SOSUI-GramN [25] , [26] , [27] . Pre-computed genome results were downloaded from the pSORTb server , which uses several analytical methods for predicting a final localization including prediction of extracellular proteins . The R . typhi predicted proteome was submitted to the SOSUI-GramN software which uses only physicochemical factors of the total sequence and the N- and C-terminal signal sequence to predict protein localizations in Gram-negative bacteria . The SecretomeP 2 . 0 server [29] was utilized to generate predictions of protein secretion not initiated by signal peptides ( i . e . non-classically secreted proteins ) . This algorithm also integrates information on post-translational and localization aspects of the protein from other feature prediction servers . CoBaltDB [24] , a comprehensive database that compiles prediction outputs from multiple sources regarding complete prokaryotic proteomes , was also utilized for comparison of subcellular localization predictions . Orthologs of the five Scas encoded with the R . typhi genome were extracted from the PATRIC web site [73] . Initially , only protein sequences from 16 completely sequenced rickettsia genomes were included: R . bellii str . RML369-C; R . bellii str . OSU 85 389; R . canadensis str . McKiel; R . typhi str . Wilmington; R . prowazekii str . Madrid E; R . prowazekii str . Rp22; R . felis str . URRWXCal2; R . akari str . Hartford; Rickettsia endosymbiont of Ixodes scapularis ( REIS ) ; R . massilae str . MTU5; R . peacockii str . Rustic; R . rickettsii str . Sheila Smith; R . rickettsii str . Iowa; R . conorii str . Malish 7; R . sibirica str . 246; R . africae str . ESF-5 . Except for split ORFs in R . prowazekii genomes , pseudogenes ( as detected using tblastn searches across all genomes ) were not included in the analyses . Sequences for each Sca orthologous group were aligned using MUSCLE v3 . 6 [74] , [75] with default parameters ( full annotated alignments are available in Supplement 1 ) . Phylogenies were estimated for Sca1 , Sca2 , Sca4 , and Sca5 under maximum likelihood using RAxML [76] ( Figure S2 ) . A gamma model of rate heterogeneity was used with estimation of the proportion of invariable sites . Branch support was evaluated from 1000 bootstrap pseudoreplications . For Sca1 , Sca2 , Sca3 , and Sca5 , the alignments were divided into the passenger domain and AT domain . Phylogenies of each domain from Sca1 , Sca2 and Sca5 were estimated . For Sca4 , sequences from additional rickettsial species ( and plasmids ) were included in a larger analysis , with alignment and phylogeny estimation as described above . Repeat regions within Sca4 and the passenger domains of Sca1 , Sca2 , Sca3 , and Sca5 were predicted using HHrepID [77] . Sub-confluent monolayers of L929 cells in 6-well plates were infected with an MOI of 10 R . typhi str . Wilmington for 0 , 5 , 15 or 30 min and 1 , 8 , 24 , 48 and 120 h . Rickettsiae-infected cells were washed briefly with cold PBS and immediately disrupted with the lysis buffer , Buffer RLT , the first step in the total RNA extraction procedure using the AllPrep DNA/RNA kit ( Qiagen , Valencia , CA ) . Total RNA was DNase-treated with the RNase-Free DNase Set ( Qiagen ) then cleaned up and concentrated with the RNeasy MinElute Cleanup kit ( Qiagen ) . DNA removal was confirmed using the SuperScriptIII One-step RT-PCR with Platinum Taq kit ( Invitrogen , Carlsbad , CA ) and cDNA was made using the SuperScriptIII First-Strand Synthesis SuperMix for qRT-PCR kit ( Invitrogen ) . Expression of the Sca genes was analyzed as previously described [78] with some modifications . Briefly , gene expression was detected using LUX primers ( Invitrogen ) in a multiplex format . Primers for rpsL , GAPDH and one of the Sca genes were included in each PCR reaction for amplification using the Platinum Quantitative PCR SuperMix-UDG kit ( Invitrogen ) . Reactions were performed in duplicate on a MX3005P Stratagene real-time thermal cycler . Primer pairs were designed using the D-LUX Designer ( Invitrogen ) and chosen based on a melting temperature of 55°C and optimal amplicon size of 90–200 bp . Primer pairs used in this study are listed in Table 1 . Cycling conditions were as follows: one cycle of 50°C for 1 h; one cycle of 95°C for 10 min; and 40 cycles of 95°C for 30 s , 60°C for 1 min , and 72°C for 30 s followed by a disassociation cycle of 95°C for 1 min , a 30-s hold at 55°C , and a ramp up at 0 . 1°C/s to 95°C for a 30-s hold . Data were imported and analyzed as described in Ceraul et al 2007 . Cycle thresholds above 45 were excluded from analysis and the results from three experiments were combined and the median normalized expression values were calculated . Polyclonal antibodies were raised against predicted immunogenic peptides from each of the Sca proteins . Complete protein sequences for Sca proteins 1–4 ( accession numbers YP_066986 , YP_067021 , YP_067397 and YP_067439 respectively ) were submitted to Custom Peptide/Antibody Services ( Invitrogen ) for peptide design based on antigenicity , residue accessibility and hydrophilicity predictions . The following peptide sequences were chosen for use in the custom antibody PolyQuik rabbit protocol ( Invitrogen ) : Sca1 - 753NYNKGEKNYDSDFK767 ( Sca1753–767 ) , Sca2 – 496LNNQNVQDENNKEW509 ( Sca2496–509 ) , Sca3 – 314IKGINNEEERLNLK327 ( Sca3314–327 ) , Sca4 – 263HYEEGPNGKPQLKE276 ( Sca4263–276 ) . All peptides were within the predicted passenger domains of the proteins and were conjugated to form multiple antigen peptides ( MAPs ) for enhanced immunogenicity . For Sca5 , the peptide sequence 651NDGSVHLTHNTYLI665 ( Sca5651–665 ) was chosen based on the immunogenicity of the R . prowazekii and R . conorii Sca5 homologs [79] , [80] . The Sca5 peptide was also conjugated to form a MAP and used in the Premium Rabbit Protocol ( Invitrogen ) . For all downstream applications , polyclonal rabbit and R . typhi-immune rat sera were purified using the Melon Gel IgG Spin Purification Kit ( Pierce Thermo Scientific ) . The purified sera were eluted at a 10-fold dilution and subsequent dilutions are given with respect to undiluted sera . For immunoblots to determine antibody specificity , pellets of uninfected or R . typhi infected L929 cells were solubilized in 2X LDS loading buffer ( Invitrogen ) with a reducing agent and boiled for 10 min at 100°C . Samples were run on NuPAGE 4–12% Bis-Tris gels in MOPS buffer or , when blotting for Sca3 , 3–8% Tris-Acetate gels in Tris-Acetate buffer ( Invitrogen ) then transferred to PVDF membranes which were processed following a standard immunoblotting protocol . Membranes were probed with anti-Sca sera diluted 1∶250 and developed with SuperSignal West Pico Chemiluminescent Substrate ( Pierce Thermo Scientific ) . Infections of 6-week old female laboratory white rats , Rattus norvegicus Sprague-Dawley ( Charles River Laboratories Inc . , Wilmington , MA ) were carried out under BSL3 conditions . Rats received intradermal injections of 0 . 5 mL DMEM supplemented with 15% FBS containing 1×103 R . typhi at the base of the tail . Control rats were injected with media only . Rats were euthanized on day 9 and organs were harvested , cut into pieces ( ≈0 . 5 cm3 ) and placed in fixation buffer [0 . 05 M phosphate buffer , 0 . 1 M lysine , 2 mg/mL sodium periodate , 1% paraformaldehyde ( PFA ) ] overnight . The fixation buffer was replaced with a solution of 10% sucrose in phosphate buffer and incubated for 30 minutes at 4°C with occasional shaking; this step was repeated with 20 and 30% sucrose solutions . Tissue samples were embedded in Tissue-Tek Optimal Cutting Temperature ( OCT ) Compound ( Sakura Finetek USA , Inc . , Torrance , CA ) , frozen in a liquid nitrogen/isopentane bath and stored at −80°C . Sections ( 2–3 µm ) were prepared using a Leica CM1900 Cryostat ( Leica Microsystems Inc . , Bannockburn , IL ) and stained as described below . Infections of membrane-adapted C . felis ( HESKA Corp . , Loveland , CO ) fleas were performed in a feeding apparatus under BSL3 conditions . Fifty fleas were placed into membrane feeding capsules and provided either 4 mL of uninfected or infected whole sheep's blood in a feeding reservoir . For an infection , renografin-purified R . typhi was added to blood for a concentration of 2 . 5×105 rickettsiae per mL . On day 3 , uninfected blood was added to the existing blood in all of the feeding reservoirs . Fleas were harvested at days 3 , 5 , 10 and 14 post-infection and placed in either 4% PFA or 4F1G fixative ( 4% PFA , 1% glutaraldehyde , 0 . 1 M PIPES , 0 . 1 M sucrose , 2 mM CaCl2 ) for electron microscopy overnight at 4°C . Fleas fixed in 4% PFA were embedded in Tissue-Tek OCT Compound and frozen at −80°C prior to sectioning; 3–5 µm sections were prepared using a Leica CM1900 Cryostat . OptiPrep-purified rickettsiae were incubated with mixtures of anti-R . typhi rat immune serum ( 1∶250 ) and anti-Sca rabbit serum ( 1∶50 ) with end over end mixing for 1 hour . Controls containing no primary antibodies , anti-R . typhi serum only or pre-immune rabbit serum only were also prepared . Bacteria were washed twice with 200 µl PBS then resuspended in the appropriate secondary antibody mixtures ( Alexa Fluor 488-conjugated donkey anti-rat 1∶500 , Alexa Fluor 647-conjugated donkey anti-rabbit 1∶500 , Alexa Fluor 647-conjugated donkey anti-rat 1∶500 ) and incubated for 30 min with mixing . Bacteria were washed and resuspended in 100 ul of 4% paraformaldehyde in PBS and incubated for 20 min with mixing then washed again prior to being resuspended in 500 µl PBS for flow cytometry analysis . Samples were analyzed on a BD FACSCanto II instrument ( BD Biosciences , San Jose , CA ) using the 488 nm ( to detect A488-conjugated anti-R . typhi immune serum staining ) and 633 nm ( to detect A647-conjugated anti-Sca antibody staining ) lasers . Rickettsiae stained with anti-R . typhi serum and either A488-conjugated anti-rat or A647-conjugated anti-rat secondary antibodies served as positively stained controls . Flow cytometry analyses were performed at the University of Maryland Greenbaum Cancer Center Shared Flow Cytometry Facility . For electron microscopy , 48 h infected L929 cells were washed three times with PBS then fixed for at least one hour in PFGPA . 1 fixative ( 2 . 5% formaldehyde , 0 . 1% glutaraldehyde , 0 . 03% picric acid ( trinitrophenol ) , 0 . 03% CaCl2 , 0 . 05 M cacodylate buffer pH 7 . 3–7 . 4 ) . After washing in 0 . 1 M cacodylate buffer cells were scraped off the plastic , pelleted and processed as previously described [81] . Briefly , the pellets were stained en bloc with 2% aqueous uranyl acetate , dehydrated in 50% then 75% ethanol and embedded in LR White resin medium grade ( Structure Probe , West Chester , PA ) . Ultrathin sections were cut on a Leica Reichert Ultracut S ultramicrotome and collected onto Formvar-carbon coated nickel grids ( Electron Microscoy Sciences [EMS] , Hatfield , PA ) . The grids were incubated in a wet chamber sequentially on drops of blocking buffer ( 0 . 1% BSA and 0 . 01 M glycine in 0 . 05 M Tris-buffered saline [TBS] ) , then on primary antibody with appropriate dilution in 1% BSA in 0 . 05 M TBS ( diluting buffer ) for 1 hr at room temperature and then overnight at 4°C . Primary antibodies were used at a 1∶50 dilution . After washing in blocking buffer , grids were incubated with a goat anti-rabbit IgG secondary antibody conjugated to 15 nm colloidal gold particles ( Aurion , EMS ) , diluted 1∶20 in diluting buffer for 1 hr at room temperature . After washing in TBS and distilled water grids were fixed in 2% aqueous glutaraldehyde , washed , stained with uranyl acetate and lead citrate and examined in a Philips 201 or Philips CM-100 Electron microscope at 60 kV . Surface labeling of whole rickettsiae for imaging by immuno-electron microscopy was performed as follows . 107 purified rickettsiae were washed in 1X PBS and suspended in 4F1G fixative for 15 min . Rickettsiae were washed and resuspended in 10 mM HEPES; 20 µl drops were placed on Formvar-carbon coated nickel grids . Samples were blocked with 5% BSA , 0 . 1% CWFS ( cold water fish skin ) gelatin in HEPES for 15 min at room temperature . Antiserum to each Sca was diluted in 1% BSA , 0 . 1% CWFS gelatin diluting buffer for 1 hour at room temperature and then 4°C overnight . Grids were washed three times with 1X HEPES followed by incubation with goat anti-rabbit IgG secondary antibody conjugated to 15 nm colloidal gold particles ( Aurion ) , diluted 1∶20 in diluting buffer for 1 hr at room temperature . Samples were fixed with 1% paraformaldehyde for 5 min at room temperature then washed three times with ddH2O for 5 min per wash . Finally , rickettsiae were negatively stained by incubation with 1% ammonium molybdate for 15 min at room temperature . Grids were viewed as noted above . Staining was performed at biosafety level 2 . For staining of rat tissues , antisera were directly labeled with either an Alexa Fluor 350 dye ( Sca antisera ) or Alexa Fluor 532 dye ( anti-R . typhi serum ) ( Molecular Probes ) . Prior to staining , rat sections were treated with a 100 µg/ml solution of DNase-free RNase ( Roche , Indianapolis , IN ) in 2X SSC ( 0 . 3 M NaCl , 0 . 03 M sodium citrate , pH 7 . 0 ) for 20 minutes at 37°C . Sections were washed briefly three times with 2X SSC , blocked with 5% BSA-2X SSC for 15 min then incubated with Alexa Fluor 350-conjugated rabbit anti-Sca antibody and Alexa Fluor 532-conjugated rat immune serum to whole R . typhi diluted 1∶100 and 1∶250 respectively in 2X SSC . Slides were placed in a humidifying chamber for 30 min at 37°C . Slides were mounted with VectaShield fluorescent mounting medium ( Vector Laboratories , Burlingame , CA ) for observation . For flea sections , the samples were blocked with 5% BSA-PBS then sequentially incubated with an anti-Sca serum followed by rat immune serum to whole R . typhi . Positive staining was assessed using Alexa Fluor 594 donkey anti-rabbit IgG and Alexa Fluor 488 donkey anti-rat IgG secondary antibodies ( Molecular Probes ) each diluted 1∶500 in 1% BSA-PBS for 30 min at 37°C . Slides were mounted with VectaShield fluorescent mounting medium with DAPI ( Vector Laboratories ) . YP_067439: Rickettsia typhi str . Wilmington; ADE30028: Rickettsia prowazekii Rp22 , AF163010_1: Rickettsia sp . IRS 4; AF155056_1: Rickettsia sp . Bar29; AAZ83584: Rickettsia asiatica; YP_001499380: Rickettsia massiliae MTU5; AAZ95593: Rickettsia tamurae; Q9AJ81: Rickettsia rhipicephali; Q9AJ79: Rickettsia japonica YH; YP_002916099: Rickettsia peacockii str . Rustic; AEK74699: Rickettsia heilongjiangensis 054; ABQ02467: Rickettsia sp . IG-1; AF163004_1: Rickettsia honei; ABD34821: Rickettsia raoultii; ACT33310: Candidatus Rickettsia tasmanensis; AF163009_1: Rickettsia Helvetica; ADH15759: Rickettsia aeschlimannii; NP_360304: Rickettsia conorii str . Malish 7; YP_002845259: Rickettsia africae ESF-5; ABQ02470: Rickettsia sp . TwKM01; ZP_00141907: Rickettsia sibirica 246; Q9AJ80: Rickettsia slovaca;AF163007_1: Rickettsia sp . A-167; YP_001650046: Rickettsia rickettsii str . Iowa; YP_001494783: Rickettsia rickettsii str . ‘Sheila Smith’; AF155058_1: Israeli tick typhus rickettsia; Q9AJ75: Rickettsia parkeri; AF163002_1: Rickettsia montanensis; AAZ78251: Rickettsia mongolotimonae; AF163001_1: Rickettsia sp . S; AAP92486: Rickettsia sp . BJ-90; YP_246741: Rickettsia felis URRWXCal2; YP_001493505: Rickettsia akari str . Hartford; Q9AJ64: Rickettsia australis; ACF20370: Candidatus Rickettsia barbariae; ADD12071: Candidatus Rickettsia andeanae; YP_001492306: Rickettsia canadensis str . McKiel; ADV19198: Candidatus Rickettsia goldwasserii; ZP_04699447: Rickettsia endosymbiont of Ixodes scapularis; YP_537939: Rickettsia bellii RML369-C; YP_001496362: Rickettsia bellii OSU 85–389; NP_220875 , NP_220874: Rickettsia prowazekii str . Madrid E; ZP_04698207: Rickettsia endosymbiont of Ixodes scapularis; YP_002922015: Rickettsia peacockii str . Rustic; ZP_04698322: Rickettsia endosymbiont of Ixodes scapularis . The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Allergy and Infectious Diseases or the National Institutes of Health . | Rickettsia typhi , a member of the typhus group ( TG ) rickettsia , is the agent of murine or endemic typhus fever – a disease exhibiting mild to severe flu-like symptoms resulting in significant morbidity . It is maintained in a flearodent transmission cycle in urban and suburban environments . The obligate intracellular lifestyle of rickettsiae makes genetic manipulation difficult and impedes progress towards identification of virulence factors . All five Scas were detected on the surface of R . . typhi using a combination of a biotin-labeled affinity assay , negative stain electron microscopy and flow cytometry . Sca proteins are members of the autotransporter ( AT ) family or type V secretion system ( TVSS ) . We employed detailed bioinformatic analyses and evaluated their transcript abundance in an in vitro infection model where sca transcripts are detected at varying levels over the course of a 5 day in vitro infection . We also observe expression of selected Sca proteins during infection of fleas and rats . Our study provides a proteomic analysis of the bacterial surface and an initial characterization of the Sca family as it exists in R . typhi . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"gram",
"negative",
"emerging",
"infectious",
"diseases",
"biology",
"microbiology",
"bacterial",
"pathogens"
]
| 2012 | Surface Proteome Analysis and Characterization of Surface Cell Antigen (Sca) or Autotransporter Family of Rickettsia typhi |
Respiratory syncytial virus ( RSV ) causes severe lower respiratory tract infections , yet no vaccines or effective therapeutics are available . ALS-8176 is a first-in-class nucleoside analog prodrug effective in RSV-infected adult volunteers , and currently under evaluation in hospitalized infants . Here , we report the mechanism of inhibition and selectivity of ALS-8176 and its parent ALS-8112 . ALS-8176 inhibited RSV replication in non-human primates , while ALS-8112 inhibited all strains of RSV in vitro and was specific for paramyxoviruses and rhabdoviruses . The antiviral effect of ALS-8112 was mediated by the intracellular formation of its 5'-triphosphate metabolite ( ALS-8112-TP ) inhibiting the viral RNA polymerase . ALS-8112 selected for resistance-associated mutations within the region of the L gene of RSV encoding the RNA polymerase . In biochemical assays , ALS-8112-TP was efficiently recognized by the recombinant RSV polymerase complex , causing chain termination of RNA synthesis . ALS-8112-TP did not inhibit polymerases from host or viruses unrelated to RSV such as hepatitis C virus ( HCV ) , whereas structurally related molecules displayed dual RSV/HCV inhibition . The combination of molecular modeling and enzymatic analysis showed that both the 2'F and the 4'ClCH2 groups contributed to the selectivity of ALS-8112-TP . The lack of antiviral effect of ALS-8112-TP against HCV polymerase was caused by Asn291 that is well-conserved within positive-strand RNA viruses . This represents the first comparative study employing recombinant RSV and HCV polymerases to define the selectivity of clinically relevant nucleotide analogs . Understanding nucleotide selectivity towards distant viral RNA polymerases could not only be used to repurpose existing drugs against new viral infections , but also to design novel molecules .
Respiratory syncytial virus ( RSV ) is a non-segmented , single-stranded , negative sense RNA virus and a member of the family Paramyxoviridae , which also includes human metapneumovirus and parainfluenza virus type-3 ( PIV-3 ) . RSV infection and resulting clinical sequelae usually last 1–2 weeks and results in mild “cold-like” symptoms in the majority of adults . However , RSV is an important pathogen in the elderly , immunocompromised patients , and patients with cardiopulmonary disease [1 , 2] . RSV is also a leading cause of lower respiratory disease in infants [3 , 4] . In 2005 , an estimated 33 . 8 million episodes of RSV infection occurred worldwide in infants and young children with most of these occurring in otherwise healthy full-term infants . Of these , at least 3 . 4 million severe cases of lower respiratory tract infection ( LRI ) required hospitalization , and an estimated 66 , 000 to 199 , 000 deaths occurred , mostly in the developing world [3] . In addition to the acute morbidity and mortality associated with RSV infection , LRI due to RSV may have long-term consequences as it has also been strongly associated with and implicated as a cause of childhood asthma [5] . Risk factors for severe illness associated with RSV infection include prematurity ( ≤ 35 weeks gestation ) and younger age ( under 2 years ) [6] , pulmonary deficiencies , congenital heart disease , immunosuppression , low birth weight , large family size , exposure to passive smoke , and lack of breast feeding [7] . No vaccines are available for the prevention of RSV infection . Palivizumab , a monoclonal antibody directed against RSV , is approved in the United States as a prophylactic for the prevention of serious lower respiratory tract disease caused by RSV in children at high risk of RSV . However , prophylaxis is only effective in preventing hospitalization in approximately 50% of individuals and the cost is prohibitive for otherwise healthy infants and for children in developing countries . Treatment of infants with severe RSV bronchiolitis is limited to supportive oxygen therapy and fluids . Aerosolized ribavirin , a base-modified guanosine nucleoside analog and broad-spectrum antiviral agent , is approved for hospitalized infants and young children with severe LRIs , but its use is limited due to uncertain efficacy and complexity of administration . Despite a clear unmet medical need , only a few therapeutic agents have reached clinical development [8] . These agents are primarily viral fusion inhibitors that , like neutralizing antibodies , may prevent the spread of infection from already infected cells but have no effect on intracellular viral replication [9] . A recent proof-of-concept study demonstrated that in healthy adults challenged with RSV , treatment with the fusion inhibitor , GS-5806 , reduced the viral load and the severity of clinical disease [10] . Although this study represents an important step forward in the development of RSV therapeutics , mechanisms of action involving an extracellular step may have intrinsic limitations for therapeutic applications against acute infections due to a potentially reduced window of intervention . Another potential limitation to the use of fusion inhibitors is the rapid emergence of drug-selected resistance-associated variants yielding pathogenic RSV variants that are cross-resistant to this class of molecules [11] . For these reasons , and due to the scarcity of clinical data from therapeutic candidates used against RSV , the possibility of effectively treating naturally occurring RSV infection with a small molecule remains an unanswered question . With over 25 drugs approved for the treatment of serious viral diseases , nucleoside and nucleotide analogs represent the largest class of antiviral drugs . As most of these molecules were developed to treat DNA viruses such as herpesvirus , HIV , or HBV , nucleoside analogs are generally considered to be most beneficial in suppressing chronic infections . Antiviral nucleosides such as zidovudine ( AZT ) or lamivudine ( 3TC ) share the same mechanism of action—inhibition of viral DNA polymerases through their triphosphate form , causing chain termination after incorporation of the nucleotide to the DNA [12] . The hepatitis C virus ( HCV ) inhibitor sofosbuvir is the first FDA-approved antiviral of this class to inhibit a viral RNA polymerase . Unfortunately , there are currently no examples of ribonucleoside analogs that are clinically efficacious against RNA viruses other than HCV [13] . Favipiravir ( T-705 ) and BCX4430 are two broad-spectrum nucleoside analogs currently under evaluation in Ebola-infected patients based on encouraging animal efficacy [14–16] , but their precise mechanism of action and degree of selectivity towards viral and host targets remain unclear [16 , 17] . The small molecule ALS-8176 ( also referred to as ALS-008176 ) is a first-in-class nucleoside analog prodrug currently undergoing clinical evaluation for the treatment of RSV infection in hospitalized infants ( ClinicalTrials . gov identifier: NCT02202356 ) . ALS-8176 and its parent cytidine analog ALS-8112 were recently discovered to be potent inhibitors of RSV replication in vitro [18] . In clinical trials , ALS-8176 given orally was efficacious against RSV infection in adult volunteers [19] . In this paper , we present the mechanism of action and the antiviral selectivity of ALS-8176 and ALS-8112 . We show that ALS-8176 inhibited RSV replication in non-human primates while ALS-8112 inhibits a broad panel of RSV A and B subtypes in vitro , as well as RSV-related viruses from the Paramyxoviridae and the Rhabdoviridae families . We identify the RNA polymerase function of the L protein of RSV as the molecular target of ALS-8112 by selecting and characterizing drug resistance-associated mutations . In enzymatic assays , we show that the 5'-triphosphate form of ALS-8112 ( ALS-8112-TP ) causes immediate chain termination of RNA synthesis and inhibition of the viral polymerase activity , a hallmark of many approved antiviral nucleoside analogs . Finally , we provide a mechanistic basis for target selectivity by evaluating clinically-relevant ribonucleotide analogs that specifically inhibit the RNA polymerase of RSV , HCV , or both . We find that subtle structural changes in nucleotides dramatically alter their antiviral spectrum . The potential medical implication of these findings is discussed .
A series of ribonucleoside analogs was recently identified as inhibiting the replication of RSV , and the optimization of the precursor molecules led to the chemical synthesis of 2'-fluoro-4'-chloromethyl ( 2'F-4'ClCH2 ) cytidine , referred to as ALS-8112 ( Fig 1A ) [18] . The nucleoside analog ALS-8112 did not significantly decrease the viability of human epithelial ( HEp-2 ) cells after 5 days ( with the concentration resulting in 50% cytotoxicity [CC50] > 100 μM ) ( Fig 1B ) . Using the same cell type and assay duration , ALS-8112 inhibited the RNA replication of RSV A2 and B1 strains with concentration resulting in 50% inhibition ( EC50 ) values of 0 . 153 ± 0 . 076 μM , and 0 . 132 ± 0 . 055 μM , respectively ( Fig 1B ) . In addition , ALS-8112 demonstrated potent inhibition of a range of diverse RSV clinical isolates with comparable EC50 values ( Table A in S1 Text ) . To understand the role of 5'-triphosphate formation in antiviral effect , we synthesized ALS-8112-I , an analog of ALS-8112 in which the 5'-hydroxyl group was replaced by iodine ( Fig AA in S1 Text ) . Because of this modification , ALS-8112-I cannot form any triphosphate in vitro . As expected , ALS-8112-I did not significantly inhibit the luciferase activity in the RSV replicon ( Fig AB in S1 Text ) . The in vitro antiviral activity of ALS-8112 was also characterized in a three-dimensional tissue culture system . This system consists of normal , human-derived tracheal/bronchial epithelial cells cultured to form a pseudo-stratified cell arrangement closely resembling the epithelial tissue of the respiratory tract [20 , 21] . The apical surface of the cultures contains numerous microvilli and cilia and the presence of tight junctions resembles the normal epithelial tissue of the lung . In this in vitro three-dimensional lung model , ALS-8112 was added to the basal media and incubated overnight before RSV strain A2 was added to the apical side of the system . The antiviral activity of ALS-8112 in human donor cells ( n = 3 ) is described in Fig 1C and Fig B in S1 Text . In the donor cells , ALS-8112 inhibited RSV RNA replication with an EC50 ranging between 0 . 09 and 0 . 73 μM , and 90% inhibition ( EC90 ) between 1 . 3 and 2 . 7 μM . We conclude that ALS-8112 is a pan-strain inhibitor of RSV replication in vitro , and that the antiviral activity of ALS-8112 is dependent upon the formation of its 5'-triphosphate metabolite . We have previously shown that ALS-8112 and ALS-8176 inhibit RSV replication in a subgenomic luciferase-based replicon assay [18] . This effect also correlated with changes in viral protein synthesis , as judged by the reporter red fluorescent protein produced by recombinant RSV infectious particles ( Fig 1D ) . Because of its high oral bioavailability [18] , the 2' , 3’-diester prodrug ALS-8176 was evaluated for in vivo efficacy in non-human primates . The effect of oral treatment with ALS-8176 on the replication of RSV A2 in the African Green monkey model was studied . ALS-8176 was administered twice daily ( BID ) , starting with a single loading dose of 200 mg/kg one day prior to RSV inoculation and followed by maintenance doses of 50 mg/kg given BID for a total treatment duration of 6 days . Based on the pharmacokinetic profile of ALS-8176 [18] , this dose was predicted to deliver an intracellular level of ALS-8112-TP in the lung of ~5-fold the level required to achieve the antiviral EC90 in vitro . At the end of treatment , RSV RNA titers reached approximately 1 × 106 copies/mL in bronchoalveolar lavage ( BAL ) samples from the four animals dosed with vehicle . In contrast , in all four ALS-8176-treated animals , RSV RNA was undetectable ( < 50 copies/mL ) in samples collected ( Fig 1E ) . This represents a difference of > 4 log10 copies/mL between the RSV RNA titers of the vehicle-treated animals and those of the ALS-8176-treated animals indicating that ALS-8176 significantly suppressed RSV replication in vivo . A similar profile was observed in nasopharyngeal ( NP ) swab samples ( Fig 1E ) , although in this case the difference in RSV RNA titers between the vehicle and ALS-8176-dosed groups was about 3 log10 copies/mL . These results demonstrate that ALS-8176 significantly inhibits RSV replication in vivo . RSV A2 viruses were repeatedly passaged in HEp-2 cells with or without increasing concentrations of ALS-8112 . Two independent adaptations were conducted , as well as a no-drug control virus pool that was passaged at the same time . After > 35 passages , the virus grown in the presence of ALS-8112 from each ALS-8112-treated pool exhibited a > 50-fold EC50 shift in the antiviral assay when compared with the control virus pool cultured for the same duration in the absence of ALS-8112 . When the complete RSV genome was sequenced , all ALS-8112-selected viruses showed four amino acid substitutions ( QUAD ) ; methionine 628 to leucine ( M628L ) , alanine 789 to valine ( A789V ) , leucine 795 to isoleucine ( L795I ) , and isoleucine 796 to valine ( I796V ) within the CRIII of the RdRp , the RNA polymerase coding region of the RSV L gene ( Fig 2A ) . Within the ALS-8112 drug selection pool , >95% of the viruses carried all four mutations on the same genome . Three of the four amino acid changes associated with ALS-8112 resistance are co-located within motif B . This particular sequence of the CRIII region is in close proximity to the catalytic motif C residues ( 810-GDNQ-813 ) that are responsible for nucleotide incorporation by paramyxovirus RNA polymerases . No mutations were identified in any other RSV genes . Phenotypic validation of the QUAD mutations was conducted by reverse genetics using the RSV minigenome system . In this assay , four plasmids encoding RSV N , P , M2-1 , and L proteins were co-transfected into HEp-2 cells . The transient expression of each of the corresponding proteins leads to the formation of a functional RSV polymerase complex , and its RNA-dependent RNA polymerase ( RdRp ) activity was monitored in the cells using a luciferase-based reporter [22] . We mutated the L gene at all four positions previously identified from sequencing , and analyzed the effect of these substitutions on the inhibition potency of the drug . In this system , ALS-8112 inhibited the wild-type RSV polymerase-dependent luciferase activity with an EC50 of 0 . 25 ± 0 . 04 μM ( Fig CA in S1 Text ) . In comparison , ALS-8112 inhibited the QUAD-mutant RSV polymerase-dependent luciferase activity with an EC50 of 9 . 7 ± 5 . 4 μM , only reaching about 60% maximum inhibition at the top concentration ( Fig 2B and Fig CA in S1 Text ) . The presence of the four amino acid changes did not significantly affect the amount of firefly luciferase signal , which may indicate similar RSV polymerase activity between the two clones ( Fig CB in S1 Text ) . The 39-fold shift in inhibition potency caused by the four mutations in the L gene indicates that ALS-8112 inhibits RSV replication by targeting the L protein responsible for the viral RNA transcription and replication functions . A detailed description of the characterization of the ALS-8112 resistant virus and the role of each individual mutation will be the subject of a separate study . Crude ribonucleoprotein ( RNP ) complex containing the RSV L protein was extracted from virus-infected cells . The RdRp activity was specific to the viral RNP complex , and was not found in uninfected cells ( Fig 2C ) . The viral RNA transcription activity was inhibited by ALS-8112-TP , but not by ALS-8112 5'-monophosphate ( ALS-8112-MP ) , nor by the non-phosphorylated parent nucleoside , ALS-8112 . The RNA transcription activity of the RSV–RNP complex was dose-proportionally inhibited by ALS-8112-TP with an IC50 of 0 . 020 ± 0 . 008 μM ( Fig 2D and Fig CC in S1 Text ) . The same RNP extraction was performed with cells infected with the QUAD mutated virus . The RdRp activity of the QUAD RNP complex was only 57% inhibited by saturating concentrations of ALS-8112-TP , with an IC50 of 2 . 9 ± 0 . 4 μM ( Fig CD in S1 Text ) . This corresponds to a 145-fold loss in inhibition potency compared to the wild-type protein ( Fig 2E and Fig CE in S1 Text ) . The inhibition of RNA synthesis catalyzed by the wild-type RNP complex in the presence of ALS-8112-TP was inversely proportional to the concentration of cytidine triphosphate ( CTP ) in the reaction ( Fig 3A ) . However , the concentration of adenosine triphosphate ( ATP ) , guanosine triphosphate ( GTP ) , and uridine triphosphate ( UTP ) had no effect on the inhibition potency of ALS-8112-TP . Therefore , ALS-8112-TP is a competitive inhibitor of cytidine monophosphate ( CMP ) incorporation into nascent RNA product by the RSV RNP complex . To understand the effect of ALS-8112-TP on the production of RNA transcripts by crude wild-type RNP complex , products of RNA synthesis were treated with RNase H and visualized by high-resolution gel electrophoresis [23] . Using this method , individual gene transcripts were separated , and their inhibition correlated with the increase in concentration of ALS-8112-TP ( Fig 3B ) . Since ALS-8112 inhibited all tested lab-adapted and clinical strains of RSV equally well , the potential for cross-species antiviral activity was evaluated in infected cells . ALS-8112 also inhibited the replication of RSV-related PIV-3 from the Paramyxoviridae family , as well as the more distant vesicular stomatitis virus from the Rhabdoviridae family ( Table 1 , and Fig D in S1 Text ) . However , ALS-8112 did not inhibit unrelated representative viruses with negative-strand segmented RNA or with positive-strand RNA genomes such as influenza A and HCV , respectively ( Table 1 , and Fig D in S1 Text ) . This trend in pan-antiviral profile limited to non-segmented ssRNA ( - ) viruses converged with the selectivity of ALS-8112-TP towards viral RNA polymerases . In enzymatic assays , ALS-8112-TP inhibited the RdRp activity of PIV-1 polymerase , but not of RSV-unrelated polymerases from influenza A or HCV ( Table 2 , and Fig E in S1 Text ) . When tested for selectivity against human enzymes , ALS-8112-TP was not recognized as a substrate for human mitochondrial RNA polymerase ( Fig F in S1 Text , and [18] ) . In comparison , 4'N3-CTP , the triphosphate form of balapiravir , was efficiently incorporated into the RNA by the mitochondrial enzyme ( Fig F in S1 Text , and [24] ) . 2'F-CTP , but not 4'ClCH2-CTP , was also recognized as a substrate . Therefore , the discrimination of ALS-8112-TP by the mitochondrial polymerase was provided mainly by its 4'ClCH2 moiety . We conclude that ALS-8112-TP is a potent and selective chain terminator of RSV and other Paramyxoviridae RNA polymerases , with low potential for interaction with human polymerases . Recently , it has been shown that , once co-purified , recombinant RSV L and P proteins form a dimer that recognizes synthetic RNA templates to synthesize short products ( Fig 4A , [25] ) . In a similar assay format , the RNA polymerase activity of recombinant RSV L-P polymerase in complex with a short synthetic primer/template substrate was monitored ( Fig 4B ) . In the presence of GTP alone or GTP+ATP , wild-type RSV L-P RNA polymerase complex specifically extended the RNA primer by 1 or 3 bases , respectively ( Fig 4C , lanes 1 and 2 ) . As a control , the L-P protein variant containing a single N812A mutation within the catalytic site of the RdRp region of L was inactive ( lanes 3 and 4 ) . Under these conditions , no signs of cytidine mis-incorporation were observed in the absence of ATP ( lane 5 ) . However , the specific incorporation of natural CMP in the presence of GTP and ATP led to the formation of +7 full-length RNA products ( lane 7 ) . In comparison , ALS-8112-MP was also incorporated into the nascent RNA at a specific +4 position opposite to a single guanosine on the template ( lane 8 ) , but not at any other positions ( lane 6 ) . After the incorporation of ALS-8112-MP , no subsequent nucleotide could be incorporated at the 3ʹ-end of the RNA primer and full-length product formation ( +7 position ) was not achieved ( lane 8 ) . This finding demonstrates that ALS-8112-TP inhibits the polymerase activity of the RSV L-P RNA polymerase complex by immediate termination of chain synthesis . The incorporation of ALS-8112-MP by RSV polymerase was efficient , as judged by the modest 13±2 . 5-fold discrimination relative to natural CTP ( Fig 4D and 4E ) . In comparison , the specific inhibitor of HCV RNA polymerase mericitabine-TP ( 2'F-2'Me-CTP ) was not recognized as a substrate for RSV polymerase under standard assay conditions . Even at a concentration of 300 μM , the level of incorporation of mericitabine-MP was less than 20% ( Fig G in S1 Text ) . Compared to natural CTP , this represents a discrimination level greater than 5 , 000-fold . Consistently , mericitabine did not inhibit RSV replication in the subgenomic replicon assay ( Table 3 ) . To understand at the molecular level why ALS-8112-TP , unlike mericitabine-TP , is a substrate for RSV polymerase , we analyzed the incorporation profile of a series of structural intermediates between ALS-8112-TP and mericitabine-TP . Nine CTP analogs were modified to contain either an OH , F , diF , or F-Me at the 2'-position , and either an H , N3 , or ClCH2 at the 4'-position ( Fig 5A ) . Some of these molecules included the triphosphate form of other clinically relevant molecules such as gemcitabine ( 2'diF-CTP ) and balapiravir ( 4'N3-CTP ) . We found that 2'F- and 2'diF-CTP were substrates for RSV polymerase , but did not cause immediate chain termination ( Fig 5A ) . In contrast , 2'F-4'ClCH2- ( ALS-8112-TP ) , 4'N3- , 2'F-4'N3- and 2'diF-4'N3-CTP were both substrates and chain terminators of recombinant RSV polymerase . Finally , 4'ClCH2-CTP seemed less efficiently recognized by the enzyme , while 2'diF-4'ClCH2- and 2'F-2'Me-CTP ( mericitabine-TP ) were almost completely inactive . Balapiravir-TP ( 4'N3-CTP ) is a known inhibitor of HCV polymerase [26] , which suggested that other CTP analogs could have dual RSV/HCV inhibition properties . This was further investigated by measuring the IC50 values of all compounds against HCV polymerase and RSV RNP complex . As expected , 2'F- and 2'diF-CTP did not efficiently inhibit either of the two polymerases , most likely because they did not cause immediate chain termination ( Fig 5B ) [27] . In contrast , 4'N3- , 2'F-4'N3- , 2'diF-4'N3- , and 4'ClCH2-CTP were all dual RSV/HCV polymerase inhibitors . These results were in agreement with the dual RSV/HCV antiviral effect of balapiravir in subgenomic replicons ( Table 3 ) . From this series of molecules , 2'F-4'ClCH2- ( ALS-8112-TP ) was the only CTP analog to efficiently inhibit RSV but not HCV polymerase . We conclude that , at the structural level , the selectivity of ALS-8112 towards RSV is provided by a combination of the 2'F- and the 4'ClCH2 moieties on the ribose group ( Fig 6A and Fig H in S1 Text ) . On the other hand , the presence of the 2'Me-moiety led to a lack of recognition by RSV polymerase ( Figs 5 and 6A ) . To rationalize how the 2'F- and the 4'ClCH2 moieties contribute to the selectivity of ALS-8112-TP , we docked analogs of cytidine diphosphate ( CDP ) into the active site of the HCV polymerase ternary complex [28] ( Fig 6B ) . Three binding metrics were computed for each docking pose: 1 ) RMSD ( Å ) to the natural CDP ligand; 2 ) number of hydrogen bonds formed with the crystallographically resolved water molecule anticipated to catalyze the formation of the new phosphodiester bond to the incoming nucleotide; and 3 ) number of hydrogen bonds formed with the RNA template ( Fig IA in S1 Text ) . 2'F-2'Me-CDP had the lowest RMSD ( Fig IA in S1 Text ) and adopted a conformation comparable to CDP ( Fig 6C ) . In comparison , 2'F-4'ClCH2-CDP ( ALS-8112-DP ) had the highest RMSD and the lowest number of hydrogen bonds ( Fig IA in S1 Text ) . This unfavorable geometry was the result of a steric hindrance between the bulky 4'ClCH2 group and the side chain of Asn291 ( Fig 6D and Fig IB in S1 Text ) , also placing a hydrophobic substitution in the proximity of a polar protein residue . In contrast , the more linear and charged 4'N3 group better occupied the binding pocket near the polar Asn291 ( Fig 6E and Fig IB in S1 Text ) . The four ALS-8112-selected amino acid mutations ( QUAD: M628L , A789V , L795I , and I796V ) were engineered into the RdRp region of the RSV L gene , and the recombinant L-P protein complex was produced . The wild-type enzyme and the QUAD mutant displayed a similar level of RdRp activity ( Fig J in S1 Text ) . Using the primer extension assay described in Fig 4 , the relative incorporation efficiencies between ALS-8112-TP substrate and natural CTP by RSV polymerase QUAD mutant led to a discrimination level of 61±1 . 1-fold ( Fig 7A ) . Compared to the wild-type L-P enzyme ( 13±2 . 5-fold , Fig 4E ) , this represents a 4 . 6-fold increase in discrimination for ALS-8112-TP ( Fig 7B ) . In order to understand at the molecular level which components of ALS-8112-TP contributed to the increased discrimination by the QUAD mutant , the same incorporation efficiency experiments were repeated with 2'F-CTP . Both WT RSV polymerase and the QUAD mutant efficiently recognized 2'F-CTP , which translated to low and comparable discrimination levels relative to CTP ( Fig 7C and 7D ) . In contrast , 4'ClCH2-CTP was discriminated 288±48-fold by WT RSV polymerase , and 6 , 990±622-fold by the QUAD mutant ( Fig 7E ) . This represents an overall resistance level of 24-fold ( Fig 7F ) . Taken together , these results show that the increased discrimination of ALS-8112-TP by RSV polymerase QUAD mutant is conferred solely by the 4'ClCH2 moiety . Given the position of the last three of the four mutations within the RdRp domain , it is likely that motif B plays a critical role in the recognition by RSV polymerase of 4'-substitutions in nucleotide analogs .
In this study , we report the first detailed molecular characterization of ALS-8176 , a ribonucleoside analog that inhibits RSV replication in vitro and is efficacious in RSV-infected adult volunteers [18 , 19] . We provide data demonstrating that ALS-8176 was efficacious in RSV-infected African Green monkeys ( Fig 1E ) , and that the parent nucleoside , ALS-8112 , is a broad-spectrum yet selective inhibitor of RNA viruses within the Paramyxoviridae family , which includes several other human pathogens such as the parainfluenza viruses ( Table 1 , Fig 1B , and Fig D in S1 Text ) . ALS-8112 also inhibited the replication of vesicular stomatitis virus , a rhabdovirus . However , ALS-8112 did not inhibit RNA viruses distant from the Paramyxoviridae family , such as influenza ( segmented ssRNA- ) or HCV ( ssRNA+ ) ( Table 1 ) . This antiviral profile is consistent with the mode of action of ALS-8112 specifically targeting the RdRp domain of the viral L protein , which is conserved within the Mononegavirales [29–31] . In viruses from this order , the L protein mediates all the enzymatic functions required for genomic replication and transcription , including the RdRp activity . Our mechanistic studies provide several independent lines of evidence that ALS-8112 specifically targets the RdRp function of RSV polymerase . Viruses continuously passaged in the presence of ALS-8112 in vitro eventually developed drug resistance-associated mutations ( Fig 2A ) . Three of the four amino acid changes associated with ALS-8112 resistance are positioned within motif B , in the vicinity of the CRIII region that contains the conserved motif C residues 810-GDNQ-813 responsible for nucleotide incorporation by paramyxovirus RNA polymerases [32 , 33] . These mutations were introduced into the RSV L gene to measure the viral polymerase activity with a cell-based minigenome assay . In this assay , the presence of the four mutations in the RSV polymerase gene was associated with a 39-fold loss in antiviral potency , without any significant change in raw luciferase signal ( Fig 2B , Fig CA and CB in S1 Text ) . In order to demonstrate that the inhibition of RSV replication by ALS-8112 is mediated by its 5'-triphosphate metabolite , we synthesized the 5'-iodo derivative of ALS-8112 ( ALS-8112-I ) as a negative control . As expected , ALS-8112-I was inactive against RSV ( Fig A in S1 Text ) . The mechanism of action of ALS-8112 involving its 5'-triphosphate as the active form was further supported by the convergence in antiviral spectrum from the enzymatic assays directly showing that ALS-8112-TP inhibits the RNA polymerase of RSV and PIV-1 ( Fig 2C and Table 2 ) , but not those of influenza or HCV ( Table 2 , Fig E in S1 Text ) . Other direct evidence for the mechanism of action of ALS-8112 is provided by single nucleotide incorporation experiments using purified RSV polymerase expressed as a recombinant L-P dimer ( Fig 4A ) . In this assay , the RSV L-P protein specifically incorporated ALS-8112-MP to the growing RNA chain opposite a guanine on the template ( Fig 4B and 4C ) . The result of ALS-8112-MP incorporation was immediate and complete chain termination of RNA synthesis . This is the hallmark of many potent nucleoside/nucleotide analogs that have been used against other viruses such as HIV , HBV , and HCV ( reviewed in [12 , 34 , 35] ) . For RSV and other paramyxoviruses , ALS-8112 is the first example of an antiviral nucleoside analog that specifically targets the viral polymerase and causes chain termination . From an enzymatic standpoint , ALS-8112-TP can be considered an efficient substrate for nucleotide incorporation by RSV polymerase , with only a 13-fold discrimination relative to natural CTP ( Fig 4D and 4E ) . In the presence of the QUAD mutations , the level of ALS-8112-TP discrimination was 61-fold ( Fig 7A ) . This represents a 4 . 6-fold increase compared to the wild-type enzyme ( Fig 7B ) . This is significantly lower than the resistance phenotype measured in the RSV minigenome and in the crude RNP assays , both monitoring full genomic transcription events . Our interpretation is that the 4 . 6-fold increase in discrimination at the single nucleotide level might be amplified with the total number of cytidine incorporation possibilities occurring during complete RSV genome replication and transcription . Importantly , we found with the recombinant RSV polymerase that the resistance phenotype provided by the QUAD mutations was solely caused by the 4'ClCH2 moiety ( Fig 7C–7F ) . Given the position of the last three of the four mutations within the RdRp domain ( Fig 2A ) , it is likely that motif B plays a critical role in the recognition by RSV polymerase of 4'-substitutions in nucleotide analogs . Further experiments will be needed to fully understand the individual contribution of each of the QUAD mutations towards the resistance of RSV polymerase against ALS-8112 . Another aim of this study consisted in evaluating at the molecular level the effect of structural changes in ALS-8112-TP on the recognition by RSV polymerase . We found that the triphosphate form of the ribonucleoside analog mericitabine ( 2'F-2'Me-CTP ) once developed for the treatment of HCV infection is > 5 , 000-fold less efficiently incorporated than its natural CTP counterpart ( Fig 4D and 4E ) . Interestingly , the recombinant RSV polymerase complex incorporated other cytidine analogs known to inhibit HCV polymerase ( Fig 5A ) , such as balapiravir-TP ( 4'N3-CTP ) [26] . This led us to investigate further the effect of CTP ribose modifications on the dual inhibition of RSV and HCV polymerase . From a series of nine CTP analogs , four were able to inhibit both RSV and HCV polymerase ( Fig 5B ) . The only nucleotide selective towards RSV polymerase was ALS-8112-TP ( 2'F-4'ClCH2-CTP ) . From these experiments , we conclude that the poor enzymatic recognition of ALS-8112-TP by HCV RNA polymerase is mediated by a combination of the 2'F and the 4'ClCH2 substitutions ( Fig 6A and Fig H in S1 Text ) . Recently , Appleby et al . determined the crystal structure of the HCV RNA polymerase in ternary complex with an incoming nucleotide diphosphate ( DP ) [28] . Computer modeling of CDP analogs inside the active site of HCV polymerase revealed that the 4'ClCH2 group of ALS-8112-DP interacts unfavorably with Asn291 , whereas a 4'N3 substitution is better tolerated ( Fig 6B–6E and Fig I in S1 Text ) , in agreement with the enzyme inhibition data . The amino acid Asn291 is part of motif B , which is well conserved within plus-strand RNA virus polymerases [36] . For instance , in the crystal structure of poliovirus RNA polymerase , the equivalent Asn297 adopts a similar orientation towards the incoming nucleotide [37] , and its mutation causes significant changes in the rate of catalysis [38] . However , that part of the sequence varies in paramyxovirus and other negative-strand RNA virus polymerases [30 , 39] . Although our resistance-mutation selection study also suggests that residues in motif B may also play a role in the recognition of ALS-8112-TP by RSV polymerase , the structural basis for this observation is unknown . The rational design of RSV polymerase inhibitors is limited by the lack of a high resolution crystal structure for any paramyxovirus polymerase . Despite this limitation , the combination of molecular modeling and enzymatic analysis such as the one presented in this study should help to further optimize inhibitor potency and selectivity . Can advanced antiviral nucleosides be repurposed against neglected or emerging viral infections ? The HCV inhibitor , sofosbuvir , is the only FDA-approved ribonucleotide to specifically inhibit a viral RNA polymerase , but it is inactive against viruses distant from HCV [13] . The molecular reason why anti-HCV drugs like sofosbuvir or mericitabine are not active against negative-strand RNA viruses is an important question that has not been addressed . In this respect , we provide the first mechanistic study to measure the selectivity of clinically-relevant nucleotides by comparing their inhibition profile against the polymerase of a positive- ( HCV ) and a negative-strand ( RSV ) RNA virus . We show that minor structural modifications on the ribose of nucleotides can dramatically change their selectivity ( Figs 5 and 6 ) . From this work , further experiments will be needed to understand the true potential of ALS-8112/ALS-8176 against other non-segmented negative-strand RNA viruses related to RSV , such as filoviruses . In particular , it will be important to pinpoint the precise molecular interaction between ALS-8112-TP and the polymerase of viruses other than RSV . From a mechanistic standpoint , the study presented here is the first comparative analysis using recombinant RSV and HCV polymerases to elucidate the mechanism of action and selectivity of nucleotide analogs . Ultimately , understanding nucleotide selectivity towards distant viral RNA polymerases could not only be used to repurpose existing antivirals against previously unaddressed viral infections , but also to design novel molecules .
ALS-8176 ( also referred to as ALS-008176 ) , ALS-8112 , ALS-8112-MP , ALS-8112-TP , and all derivatives of ALS-8112-TP were synthesized at Alios BioPharma ( South San Francisco , CA , USA ) according to described procedures [18] . ALS-8112-I was synthesized as described in S1 text . The parent nucleoside ALS-8112 and its prodrug ALS-8176 were stored at 4°C in dimethyl sulfoxide ( DMSO ) , and all the phosphorylated species were reconstituted in water , aliquoted , and stored at -80°C . RSV stocks were purchased from American Type Culture Collection ( ATCC , Cat . #VR-1540 for RSV A2; VR-26 for RSV Long , and VR-1400 for RSV B1 ) . Recombinant RSV expressing the far red fluorescent protein mKate2 was previously described [40] . The viruses were amplified in Vero cells and titrated in Vero cells by the amount of virus suspension producing infection in 50% of cell cultures inoculated ( TCID50 ) . The A549 cells ( ATCC ) were maintained in Ham's F12 media with 10% fetal bovine serum [FBS] ) and 1% penicillin-streptomycin and HEp2 cells were maintained in Dulbecco's Modified Eagle medium ( DMEM ) /F-12 with 5% FBS and 1% penicillin-streptomycin . Measurement of in vitro intracellular NTP formation was performed as described in S1 text . Determination of the EC50 and CC50values of ALS-8112 in the RSV assays was performed by the following procedure . On the first day , 20 , 000 HEp-2 cells per well were plated in a 96-well plate . Each compound was serially diluted ( 1:3 ) up to 9 distinct concentrations . For the three-dimensional tissue culture system , primary NHBE cells ( EpiAirway PC-12 , MatTek Corporation ) were provided in columns embedded in agar plates [20] . On the day of their arrival , 1 . 5 mL of warm media ( AIR-100-ASY ) was added to each well of 6-well plates and then the columns were placed into the 6-well plates and incubated overnight in a 37°C humidified , 5% CO2 atmosphere . Cells were pre-incubated with compounds for 24 hours at 37°C in a 5% CO2 atmosphere . After 24 hours of pre-incubation with compounds , RSV A2 , Long , or B1 at a multiplicity of infection ( MOI ) of 0 . 5 was added to the cells , except for the background controls . The plate was then incubated for additional 4 days in the same conditions and at the end of the incubation 50 μL the supernatant from each well of the plate was collected . The RSV viral RNA was isolated from the collected supernatant of each well using a MagMAX Viral RNA Isolation Kit ( Life Technologies ) automated through the MagMAX Express-96 Deep Well Magnetic Particle Processor ( Life Technologies ) . The viral RNA was then quantified by Real Time polymerase chain reaction ( RT-PCR ) using the primer/probe set ( see sequences in S1 text ) . The RSV subgenomic replicon ( Apath , Brooklyn , NY , USA ) was used as previously described [41] . The HeLa-derived cells containing the stable RSV replicon were cultured in DMEM containing 4500 mg/L D-glucose , L-glutamine , and 110 mg/L sodium pyruvate . The medium was further supplemented with 10% ( v/v ) FBS ( Mediatech ) , 1% ( v/v ) penicillin/streptomycin ( Mediatech ) , and 10 μg/mL of Blasticidin ( BSD ) ( Invivogen ) . Cells were maintained at 37°C in a humidified 5% CO2 atmosphere . On the first day , 5000 RSV replicon cells per well were plated in a 96-well plate . On the following day , compounds to be tested were solubilized in 100% DMSO to 100 × the desired final testing concentration . Cells were incubated with compounds for 7 days at 37°C in a 5% CO2 atmosphere before measurement of the luciferase readout . Cell viability ( CC50 ) was measured with a CellTiter-Glo cell proliferation assay ( Promega ) . RSV A2 virus was passaged in increasing concentrations of ALS-8112 in duplicate , together with a parallel passage with media containing no drug as a control . After each passage , viruses from each pool were harvested , titrated by TCID50 , and stored for the next passage and further characterization . A preliminary phenotypic assessment of both treated pools after > 35 passages at a point when the ALS-8112 concentration had reached 16 μM indicated a loss in antiviral potency for ALS-8112 in the treated virus pool vs . the control virus pool . This was used as the basis to trigger RSV genome sequencing of all virus pools . Eight RSV-seronegative African Green monkeys ( Chlorocebus sabaeus ) approximately 4–8 years of age were obtained from PrimGen ( Hines , IL ) . Animals were housed at BioQual Inc . ( Rockville , MD ) under BSL2 conditions as specified by the Association for Assessment and Accreditation for Laboratory Animal Care ( AAALAC ) guidelines . All protocols were IACUC approved . The weight range of the animals varied from ~3 . 7–8 . 0 kg . Four of the animals were male . Animals were acclimated to the general housing conditions for at least 6 weeks before the study start . Prior to dose administration , sample collections or viral challenge , animals were anesthetized with ketamine hydrochloride . All ALS-8176 dose formulations were prepared fresh on each day of dosing . The dose formulation was administered via oral gavage according to standard procedures . During the ALS-8176 dosing period , animals were fasted overnight and all ALS-8176 doses were given in the fasted state . During the dosing days , animals were fed at least 1 hour after the morning dose and no later than 2 hours prior to the evening dose . ALS-8176 and vehicle were administered 24 hours prior to the viral challenge ( Day -1 ) . A modified RSV minigenome assay based on a previously described method [40 , 42] was used to assess the phenotypic shift in the quadruple mutation . In brief , HEp-2 cells were plated in 6-well plates at the density of 0 . 5 million cells/well . ALS-8112 was serially diluted and added to each of the wells and further incubated over night . On the next day , modified vaccinia virus Ankara-T7 ( MVA-T7 ) at the MOI of 1 was added to provide T7 RNA polymerase [43] . After 2 hours of viral transduction , each well was transfected with Fugene 6 ( Promega ) with 1 . 25 μg mixture of 6 plasmids including minigenome ( pGem . RSV . M5 . Luc ) , plasmids encoding human codon bias-optimized N , P , M2-1 , L genes as well as control plasmid encoding renilla luciferase pRL-SV40 . After 48 hours of further incubation , the firefly luciferase as well as renilla luciferase signals from each well were measured with dual-luciferase reporter assay system ( Promega ) . The normalized signals ( firefly luciferase over renilla luciferase ) were plotted over ALS-8112 concentrations to obtain IC50 values . Cell fractionation and extraction of crude RSV and PIV-1 RNP complexes were performed as described [23] . Unless otherwise specified , nucleoside triphosphate ( NTP ) concentrations were around their Km value: 0 . 1 μM CTP , 1 μM ATP , 2 μM UTP , and 500 μM GTP . The radioactive tracer was either 10 μCi [α-33P]rCTP ( CTP competitive mode ) or 15 μCi [α-33P]rGTP ( non-CTP competitive mode ) together with 0 . 1 μM GTP . The NTP concentration for the competition assay was either 1 or 100 μM for CTP , ATP , GTP , and UTP . The reaction was initiated with the addition of 1 . 5 μL of the virus-infected cell extract and incubated for 2 hours at 30°C in a final volume of 30 μL . Unless otherwise specified , the standard reaction was stopped by adding 25 μL of Tris-sodium-EDTA buffer containing 10 mM Tris-HCl pH 8 , 150 mM NaCl , and 100 mM EDTA . Processing of samples included elution through a G-50 spin column ( GE Healthcare ) , phenol-chloroform extraction , and ethanol precipitation . Air-dried RNA samples were reconstituted in 20 μL TBE urea gel loading buffer ( Invitrogen ) , and migrated through a 6% TBE urea PAGE gel for 1 hour at 190 V . The gel was dried and exposed to a phosphor-screen that was scanned with a Storm 860 phosphorImager ( Molecular Dynamics ) . Biochemical assays for the inhibition of influenza and HCV polymerases were performed as previously described [17 , 27] . The recombinant RSV polymerase complex was produced by co-expressing the L and P proteins of RSV in insect cells as previously described [25] . Unless otherwise specified , RNA polymerase reaction samples consisted of 0 . 2 μM of an oligonucleotide template sequence derived from the RSV leader promoter ( 5'-UUUGUUCGCGU-3' ) and 0 . 2 μM recombinant RSVL-P polymerase together with 200 μM 5'-pACGC primer , mixed in a buffer containing 20 mM Tris pH 7 . 5 , 10 mM KCl , 2 mM dithiothreitol , 0 . 5% triton , 10% DMSO , 0 . 2 U/μL RNasin ( Ambion ) , and 6 mM MgCl2 . Reactions were started at 30°C by adding specific NTPs in a final volume of 10 μL . The radioisotope tracer used for this assay was α33P-GTP . Reactions were stopped after 30 minutes by adding an equal volume of gel loading buffer ( Ambion ) . Samples were denatured at 95°C for 5 minutes , and run for 1 . 5 hours at 80 W in a 22 . 5% polyacrylamide urea sequencing gel . After the gel was dried , the product of migration was exposed to a phosphor-screen and scanned as previously described . All calculations were carried out using the Schrӧdinger software suite ( Schrӧdinger , LLC ) . A molecular model of the HCV NS5B polymerase elongation complex ( PDB 4WTC ) [28] , was generated . The protein structure was prepared with Protein Preparation Wizard . Docking grids were computed with Glide . To ensure adequate sampling of metal-chelating conformations of the diphosphate moiety during docking , ligand input conformation ensembles were generated using MacroModel , with torsional constraints in place for C4’ , C5’ and the diphosphate moiety , as observed for CDP in 4WTC . Ligand binding to the HCV NS5B polymerase catalytic site was evaluated with Glide docking calculations: SP precision level , with reference core constraints based on the CDP X-ray binding mode revealed in 4WTC ( core atoms encompassing the diphosphate moiety , C5’ , C4’ , ribose O , C1’ , and heavy atoms of the nucleobase ) . Up to 20 binding modes were collected for each ligand , without post-docking minimization . Sigmoidal dose-response curves used to generate 50% or 90% inhibitory or effective concentrations were analyzed by nonlinear regression using the four-parameter logisitic equation ( GraphPad Prism ) . All data are presented as mean ± SEM or SD as specified , with experimental uncertainties identified by error bars . Statistical significance was calculated using a 2-tailed Student's t test . Differences with a P value of less than 0 . 05 were considered statistically significant . | Viral RNA polymerase complexes mediate all of the enzymatic functions required for genomic replication and transcription in RNA viruses . Because of their essential role in the virus life cycle , RNA polymerases are major molecular targets for antiviral therapies . Sofosbuvir and mericitabine are related compounds belonging to a class of drugs called nucleoside analogs that inhibit the RNA polymerase of hepatitis C virus ( HCV ) , a positive-strand RNA virus , but have no effect on negative-strand RNA viruses . The mechanistic reason for this inactivity is unknown . The only nucleoside analog currently under clinical evaluation against respiratory syncytial virus ( RSV ) , a negative-strand RNA virus , is ALS-8176 . In this study , we present the detailed mechanism of action of ALS-8112 , the parent molecule of ALS-8176 . A multidisciplinary approach combining cellular , chemical , structural , and enzymatic methods was employed to demonstrate that the triphosphate form of ALS-8112 targets the RNA polymerase of RSV , but not of HCV . A series of molecules structurally related to ALS-8112 displayed dual RSV/HCV inhibition , whereas mericitabine only targeted HCV RNA polymerase . Understanding the molecular basis of nucleotide selectivity towards distant viral RNA polymerases could not only be used to repurpose existing drugs against new viral infections , but also to design novel molecules with broad antiviral spectrum . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| []
| 2015 | Molecular Basis for the Selective Inhibition of Respiratory Syncytial Virus RNA Polymerase by 2'-Fluoro-4'-Chloromethyl-Cytidine Triphosphate |
Candida sp . are opportunistic fungal pathogens that colonize the skin and oral cavity and , when overgrown under permissive conditions , cause inflammation and disease . Previously , we identified a central role for the NLRP3 inflammasome in regulating IL-1β production and resistance to dissemination from oral infection with Candida albicans . Here we show that mucosal expression of NLRP3 and NLRC4 is induced by Candida infection , and up-regulation of these molecules is impaired in NLRP3 and NLRC4 deficient mice . Additionally , we reveal a role for the NLRC4 inflammasome in anti-fungal defenses . NLRC4 is important for control of mucosal Candida infection and impacts inflammatory cell recruitment to infected tissues , as well as protects against systemic dissemination of infection . Deficiency in either NLRC4 or NLRP3 results in severely attenuated pro-inflammatory and antimicrobial peptide responses in the oral cavity . Using bone marrow chimeric mouse models , we show that , in contrast to NLRP3 which limits the severity of infection when present in either the hematopoietic or stromal compartments , NLRC4 plays an important role in limiting mucosal candidiasis when functioning at the level of the mucosal stroma . Collectively , these studies reveal the tissue specific roles of the NLRP3 and NLRC4 inflammasome in innate immune responses against mucosal Candida infection .
Candida sp . are dimorphic fungi that commonly colonize the oral cavity of adult humans , with overgrowth prevented by competing commensal bacteria as well as local host immune responses . Perturbations of the normal oral flora through antibiotic treatment , for example , or immunocompromised states can lead to mucosal Candida overgrowth resulting in the development of oropharyngeal candidiasis ( OPC , also known as thrush ) . Candida albicans has now been identified as the leading cause of fatal fungal infections , with mortality rates as high as 50% , and ranks 4th among all pathogens isolated from bloodstream and nosocomial infections [1]–[3] . Host recognition of Candida requires engagement of surface receptors on innate immune cells , including TLR2 and Dectin-1 [4]–[7] . A major consequence of receptor activation is the induction of pro-inflammatory gene expression including interleukin 1 beta ( IL-1β ) , a zymogen which requires proteolytic processing by caspase-1 to become biologically active [8]–[11] . Activation of caspase-1 requires signaling through recently described protein complexes termed inflammasomes , consisting of either NOD-like receptor ( NLR ) molecules or the PYHIN protein , Absent in melanoma-2 ( AIM2 ) [12]–[16] . NLRs are characterized by the presence of a Leucine Rich Repeat domain , a central NACHT domain involved in oligomerization and protein-protein interactions , and a CARD or PYRIN domain [17] . Conformational changes in NLR proteins , resulting from the introduction of activating stimuli , cause oligomerization of NLR proteins together with ASC adapters , permitting autocatalytic cleavage of pro-caspase-1 to an active state capable of cleaving pro-IL-1β . Although intracellular danger signals and crystalline compounds such as uric acid crystals , cholesterol crystals , amyloid and asbestos have been shown to activate the NLRP3 inflammasome [18]–[22] , the precise mechanism ( s ) underlying inflammasome activation are not defined . Currently , several theories have been proposed for the molecular mechanisms underlying activation of the NLRP3 inflammasome including mitochondrial ROS production [23] , phagosomal or endosomal rupture and cell membrane disturbances [24]–[27] . The NLRP3 inflammasome has been linked to IL-1β responses to pathogen-derived molecules including bacterial muramyl dipeptide [28] and toxins [20] , [28] , as well as in response to a range of bacterial , viral and fungal pathogens , including Candida albicans [6] , [29] . Another NLR molecule , NLRC4 , also forms an inflammasome capable of activating caspase-1 and IL-1β cleavage . During some bacterial infections , such as with Shigella , Salmonella , Pseudomonas or Legionella , NLRC4 detects inadvertently translocated flagellin or PrgJ rod protein , a component of the type III secretion system [30]–[35] . Although limited in vitro studies using NLRC4 deficient macrophages or dendritic cells challenged with Candida albicans revealed no defects in caspase-1-dependent IL-1β responses [29] , [36] , [37] , the role of NLRC4 in live fungal infection models has not been thoroughly defined . In this study , we sought to examine the role of other inflammasome in anti-fungal defenses in vivo . We show that infection with Candida albicans leads to up-regulation of NLRP3 and NLRC4 expression in the oral mucosa and this induction is impaired in both NLRP3 and NLRC4 deficient mice . Additionally , we reveal a role for the NLRC4 inflammasome in regulating resistance to mucosal infection with Candida as well as preventing systemic dissemination . We show that inflammasome driven IL-1β responses via both the NLRC4 and NLRP3 inflammasome are essential for epithelial antimicrobial peptide production , and other inflammatory responses including IL-18 and IL-17 in response to Candida infection . Inflammatory cell recruitment to Candida infected oral mucosa is significantly impaired in NLRC4 deficient mice compared to wild-type mice . Using bone marrow chimera mice , we reveal that the activity of NLRC4 is mediated at the level of the mucosal stroma , in contrast to that observed with NLRP3 which is active in both hematopoietic and stromal compartments . Collectively our studies show that , in addition to the NLRP3 inflammasome , there is a tissue specific role for the NLRC4 inflammasome in host sensing and immune defense to non-bacterial pathogens such as Candida albicans .
During Candida infection , the oral mucosa acts as a physical barrier to infection as well as the initial tissue to respond to fungal growth and invasion . To assess the impact of Candida treatment on the oral mucosa , we monitored gene expression levels in oral mucosa by quantitative real-time PCR . We first examined the level of NLR expression in buccal tissues of Candida infected mice and observed a strong induction of NLRP3 in wild-type mice following oral challenge with Candida albicans . Induction of NLRP3 was significantly reduced in both Nlrc4−/− and Asc−/− mice ( Figure 1A ) . Similarly , NLRC4 was induced in WT mice and negligible in Nlrp3−/− and Asc−/− mice ( Figure 1B ) . As expected , ASC was not induced in any of the strains after Candida infection . These data indicate that genetic knockdown of a single NLR may have profound effects on the expression profile of other NLR proteins and is , to our knowledge , the first evidence of cross-regulation of NLRP3 and NLRC4 . We next assessed the impact of the NLRP3 and NLRC4 inflammasomes on expression levels of members of the IL-1 family . There was a significant difference in the induction of IL-1β between the WT and Nlrc4−/− , Nlrp3−/− , and Asc−/− mice ( Figure 1C ) . This defect in IL-1β production was confirmed in the serum of infected mice at 3 days of infection ( Figure 1D ) . Levels of IL-1R1 expression were similar between WT and Nlrc4−/− or Asc−/− mice with reduced induction observed in Nlrp3−/− mice , although this was not significant ( Figure 1E ) . Induction of IL-1R antagonist ( IL-1Rn ) was not significantly different between any of the inflammasome knockout mice and WT mice ( Figure 1F ) . Overall , the induction of IL-1R1 and IL-1Rn was minimal compared to IL-1β in all the infected mice . Our previous studies demonstrated that NLRP3 signaling is critical for the prevention of fungal growth as well as dissemination in a murine model of oropharyngeal candidiasis [6] . A role for the NLRC4 inflammasome in response to oral fungal challenge has yet to be characterized . In order to ascertain the impact of loss of NLRC4 function on disease progression , we infected wild-type ( WT ) and Nlrc4−/− mice with Candida albicans as previously described [6] . Oral fungal burdens were elevated in Nlrc4−/− mice compared to WT mice by day 7 , and persistently higher fungal burdens were observed to day 21 ( Figure 2A ) . In our model of persistent , low virulence oral candidiasis , WT mice rarely show blood borne dissemination of infection , as measured by quantitative fungal burdens in the kidneys ( Figure 2B ) . In contrast , Nlrc4−/− mice show a significantly increased susceptibility to dissemination of infection , peaking at day 7 but returning to WT levels by day 21 . In agreement with these findings , Nlrc4−/− mice also had elevated gross clinical scores , a qualitative measure of oral infection severity , at all time points ( Figure 2C ) . Survival in the Nlrc4−/− mice was reduced when compared to WT mice when infected with a virulent strain of Candida albicans ( Figure 2D ) . Elevated quantitative fungal colonization was observed in tissues of the gastrointestinal tract including esophagus , stomach , and small intestine in Nlrc4−/− mice compared to WT ( Figure S1 ) . These data contrast with studies in Nlrp3−/− mice , in which it was determined that oral fungal colonization was similar at day 3 , becoming slightly elevated at day 7 and 14 , and returning to WT levels by day 21 ( Figure 2A ) . These mice exhibited elevated levels of systemic dissemination throughout the 21 day timecourse ( Figure 2B ) . By day 21 , the gross clinical score of both Nlrp3−/− and WT mice were between 0 and 1 , indicating minimal signs of infection , which contrasts the sustained elevated clinical score seen in Nlrc4−/− mice ( Figure 2C ) . Taken together , our studies imply that NLRC4 and NLRP3 are differentially functioning in the innate response to Candida infection , with NLRC4 playing a more prominent role in the clearance of oral infection . One of the earliest inflammatory cells that migrate to the site of microbial infection are neutrophils , and this chemotaxis is necessary for proper inflammatory responses and anti-microbial defenses . Given the known capacity for IL-1β to mediate leukocyte infiltration into infected tissues , we used histology to examine the impact of inflammasome deficiency on cellular infiltration to the mucosa of the tongue . By day 2 , a robust cellular infiltration was observed in the dorsal epithelium of a WT tongue , particularly in areas showing the presence of fungal hyphae and epithelial erosion ( Figure 3A ) . These cells morphologically appear to have multi-lobulated nuclei , consistent with neutrophils . In contrast , minimal cellular infiltration was observed in a tongue from Nlrc4−/− mouse , despite the presence of erosive lesions and fungal hyphae ( Figure 3C ) . Nlrp3−/− and Asc−/− tongues exhibited cellular infiltration , although not to the extent of WT; and these areas of concentrated cellular infiltrates also correlated with the presence of fungal hyphae and tissue erosion ( Figure 3E , G ) . Neutrophils have been implicated in the control of a range of microbial infections , including Candida [38]–[40] . Given the presence of significant cellular infiltration at 2d post infection , we sought to specifically characterize the extent of neutrophil infiltration in these tissues . Using a monoclonal antibody shown to specifically stain neutrophils , we observed significant neutrophil staining in the outer epithelium of the WT tongue . This immunofluorescent staining localized to the regions of increased cellularity observed in the epithelium with PAS/H staining ( Figure 3B ) . Neutrophils were also observed throughout the sub-mucosal tissue . In agreement with our finding with PAS/H staining , neutrophil influx into the Nlrc4−/− was drastically reduced ( Figure 3D ) . As expected , a significant influx of neutrophils was observed in both Nlrp3−/− and Asc−/− tongues ( Figure 3F , H ) . Intriguingly , it was observed that not only was there a reduction in neutrophil infiltration in the Nlrc4−/− tongue but the neutrophils present failed to infiltrate the epithelium where the presence of hyphae was detected . This is evidenced by a ∼25 fold reduction in the percent of dorsal epithelium that stains positive for neutrophils in Nlrc4−/− mice when compared to WT ( Figure 4 ) . Nlrp3−/− and Asc−/− mice showed a ∼3 fold decrease in positive staining ( Figure 4 ) . These findings indicate that the activation of Nlrc4−/− is required for neutrophil recruitment into infected tissues and proper trafficking to the site of active fungal infection . Recent reports have implicated the IL-17 family as a critical mediator of protective host responses to a range of extracellular pathogens , including Candida [41]–[45] . To assess the impact of inflammasome activation on IL-17 in our model of oral candidiasis , we measured expression levels of IL-17 family members in oral mucosal tissues after infection . A robust increase in IL-17A and IL-17F gene expression was detected in the oral mucosal tissue of WT animals , which was significantly reduced in Nlrc4−/− , Nlrp3−/− , and Asc−/− mice ( Figure 5A , B ) . In contrast , the induction of IL-17F was dependent on NLRP3 , but not NLRC4 or ASC ( Figure 5B ) . A robust induction of interleukin 17A receptor ( IL-17RA ) expression was detected in WT , Nlrp3−/− , and Asc−/− mice while this response was abrogated in Nlrc4−/− mice ( Figure 5C ) . As many downstream inflammatory responses are dependent on IL-17 , the failure to upregulate the IL-17A receptor may have implications for local anti-Candida inflammation and chemotaxis of inflammatory cells . We next examined expression of other inflammatory cytokines in the oral mucosa of mice infected with Candida albicans . We observed significantly lower induction of IL-18 , another cytokine requiring inflammasome mediated cleavage , in Nlrp3−/− and Asc−/− mice , while Nlrc4−/− mice showed a slight reduction compared to WT which was not statistically different ( Figure 5D ) . As shown in Figure 5E , murine CXCL1 , a homolog of human IL-8 , was dramatically induced in WT mice following Candida infection and levels were significantly reduced in all strains of inflammasome deficient mice . Induction of the pro-inflammatory cytokine IL-6 was also significantly reduced in Nlrc4−/− , Nlrp3−/− , and Asc−/− mice compared to WT ( Figure 5F ) . In addition to inflammatory cytokines responses following pathogenic encounter , immune cells in the oral mucosa , including epithelial cells , release small antimicrobial peptides designed to disrupt microbial function as well as act as strong chemoattractant signals for the migration of inflammatory cells such as neutrophils and macrophages [46]–[48] . To better define the inflammasome dependence of antimicrobial peptide responses in our murine model of OPC , we quantified several murine beta-defensins as well as cathelicidin responses in the oral mucosa following fungal infection . In contrast to murine beta-defensin 1 ( mBD1 ) , which was not induced following infection in any of the mice ( Figure 6A ) , mBD2 , -3 , -and 4 all showed elevated expression in oral tissues following Candida infection in WT mice ( Figure 6B , C , D ) . Reduced or negligible up-regulation in mBD2 and mBD4 gene expression was observed in all of the inflammasome deficient mice ( Figure 6B , D ) . In contrast , mBD3 induction was similar between WT and Nlrc4−/− mice but significantly reduced in Nlrp3−/− and Asc−/− mice ( Figure 6C ) . No appreciable induction of mBD14 , a murine homolog of human beta-defensin 3 , was observed in any of the mice ( Figure 6E ) . Another antimicrobial peptide , cathelicidin or CAMP , has also been implicated as an activator of the P2X7 receptor and a potential inducer of IL-1β release from cells [49] . Expression levels of CAMP were dramatically elevated in WT buccal tissue following Candida infection , and this induction was dependent on NLRC4 , NLRP3 and ASC ( Figure 6F ) . NLRC4 does not appear to be involved in inflammasome activation in innate immune cells exposed to Candida in vitro [29] , [36] , [37] . Yet in our in vivo model on mucosal candidiasis , NLRC4 is required for protection from mucosal colonization , prevention of early dissemination of infection , and neutrophil infiltration . Therefore , we hypothesized that the impact of NLRC4 activation during fungal infection was manifested in mucosal and/or stromal tissues versus hematopoietic cells . The innate immune system is comprised of cells of embryonic origin , including epithelial cells , as well as infiltrating leukocytes derived from the bone marrow . To assess the contribution of different inflammasome molecules in these compartments , we generated bone marrow chimera mice and infected them orally with C . albicans . Lethally irradiated recipient mice were reconstituted with bone marrow progenitor cells and allowed to fully reconstitute prior to infection . WT mice that were reconstituted with WT bone marrow exhibited no increase in oral infection or systemic dissemination of infection at 7 days when compared to non-chimeric WT mice , indicating that the chimera procedure does not predispose the mice to higher levels of fungal infection ( Figure S2 ) . WT mice reconstituted with Nlrc4−/− bone- marrow showed no significant difference in oral fungal colonization compared to WT/WT chimera mice ( Figure 7A ) . In contrast , Nlrc4−/− mice reconstituted with WT bone marrow showed enhanced oral colonization with C . albicans ( Figure 7A ) , to a degree that is similar to native Nlrc4−/− mice ( Figure 2A ) . The Nlrc4−/− mice reconstituted with WT bone marrow also exhibited increased disease severity when compared to WT/WT chimera mice ( Figure 7C ) . These results demonstrate that intact NLRC4 function in the stromal or epithelial compartment is associated with protection from mucosal infection with Candida albicans . To evaluate the role of the NLRP3 inflammasome in the stromal versus hematopoietic compartments , bone marrow chimeras were generated using Nlrp3−/− as well as Asc−/− mice . Nlrp3−/− mice receiving WT bone marrow showed difference in oral fungal burdens relative to WT/WT chimera controls ( Figure 7A ) . However , WT mice reconstituted with Nlrp3−/− bone marrow exhibited significantly elevated oral infection when compared to WT/WT chimera mice ( Figure 7A ) , although both sets of NLRP3 chimera mice showed elevated clinical scores compared to WT/WT chimera mice ( Figure 7C ) . A similar pattern was seen with ASC chimeric mice ( Figure 7A ) , where higher oral fungal colonization was observed in the WT mice receiving ASC deficient bone marrow compared to Asc−/− mice receiving WT cells or WT/WT chimera mice . These observations indicate that the function of the NLRP3/ASC inflammasome complex in infiltrating inflammatory cells is critical for control of oral mucosal infection . We next assessed the role of NLRC4 and NLRP3 in protection from systemic dissemination of infection . As a marker of dissemination , we quantified the fungal burdens of the kidneys in bone marrow chimera mice infected with C . albicans for 7 days . As shown in Figure 7B , neither Nlrc4−/− donor nor recipient chimera mice showed dissemination to the degree seen in native Nlrc4−/− mice ( Figure 2B ) , although a trend towards increased dissemination was observed compared to WT/WT chimera mice . In contrast , both Nlrp3−/− donor and recipient chimera mice showed higher systemic dissemination when compared to WT/WT chimera mice ( Figure 7B ) . WT mice receiving Asc−/− bone marrow showed enhanced dissemination of infection , whereas Asc−/− mice receiving WT bone marrow showed similar kidney fungal burdens to WT/WT chimera mice ( Figure 7B ) . These results demonstrate that the NLRP3/ASC inflammasome plays a dominant role in protection against disseminated fungal infection compared to the NLRC4 inflammasome which plays a role in protection of the host from mucosal infection .
Disseminated fungal infections present a significant health risk to both immune-competent and -compromised individuals , making studies into early host immune responses involved in the prevention of dissemination critical for the development of new therapeutic approaches . The release of inflammatory mediators from resident cells at the site of infection is critical for antimicrobial responses including the recruitment of inflammatory cells such as neutrophils and macrophages . IL-1β has been implicated in protective host immune responses to a range of infectious pathogens including viruses , bacteria , and fungi [11] . We and others have previously shown that the NLRP3 inflammasome is important for control of Candida infection in both mucosal and disseminated models [6] , [29] , [36] , [37] . Here we present the first experimental evidence implicating the NLRC4 inflammasome in the induction of protective host responses to challenge by a fungal pathogen . Our studies reveal that the NLRC4 inflammasome is important for the control of Candida albicans infection in vivo , particularly in the oral cavity , with increased oral fungal colonization and disease severity observed in Nlrc4−/− mice . Lack of NLRC4 also increased susceptibility to disseminated fungal infection , particularly early in infection . Our model of OPC using a clinical isolate recapitulates human oral infections in which mortality is rarely observed . We carried out OPC infection studies using a highly virulent strain of C . albicans ( ATCC 90234 ) and observed similar increases in oral colonization and dissemination at 7d to that observed using the oral isolate GDH2346 ( Figure S1 ) . However , the virulent strain resulted in significant mortality in Nlrc4−/− mice when compared to WT ( Figure 2E ) . When we examined cellular infiltrations in infected tongues , we observed a substantial impact in both cellular infiltration , and specifically neutrophil-influx in the absence of NLRC4 , compared to a robust neutrophil infiltration into the infected epithelium of WT mice . Interestingly , the neutrophils that were present in the Nlcr4−/− tongue remained in the sub-mucosa . This defect in neutrophil infiltration into the epithelium of the tongue may explain the extended defect in oral clearance observed in Nlrc4−/− mice throughout the course of our OPC infection studies . The importance of neutrophils in anti-fungal defenses is well known . Recent studies have shown that the innate inflammatory mileu is critical for effective neutrophil activity against fungal pathogens including IL-6 , GM-CSF , and IL-17 responses [38] , [41] , [42] , [50] , [51] . In order to better define the molecular basis for protection from mucosal infection , we evaluated a panel of innate inflammatory responses in the oral mucosa from Candida infected mice . Previous studies have demonstrated that NLRP3 expression is inducible following infection and implicate this as the rate limiting step in inflammasome activation [52] , [53] but little is known about the regulation of NLRC4 expression . IL-1β has been shown to increase the expression level of other pro-inflammatory cytokines following IL-1 receptor engagement [54]-[56] , and IL-1 receptor activation may result in the increased expression of inflammasome proteins in the responding cells , priming them for quick activation . We show that Candida infection up-regulates NLRP3 and NLRC4 expression in mucosal tissues compared to mock infected mice . Additionally , we demonstrate that this induction is impaired in Nlrp3−/− , Asc−/− and Nlrp4−/− mice . As expected , ASC expression was found to be constitutive and not induced by Candida infection ( data not shown ) . Our data provides novel insight into transcriptional changes induced by activation of inflammasome complex by an infectious pathogen . The impairment of mucosal NLRP3 induction in NLRC4 deficient mice may enhance the impact of the lack of this receptor on susceptibility to mucosal infection . As expected , we found that IL-1β was up-regulated in wild-type mice in response to fungal infection . However , these responses were partially abrogated in NLRC4 deficient mice and absent in NLRP3 and ASC deficient mice . Similarly , the IL-1 receptor 1 ( IL1R1 ) and IL-1 receptor antagonist ( IL1Rn ) were poorly induced by Candida infection in WT and inflammasome deficient mice . Consistent with published studies , we did not observe a role for the NLRC4 inflammasome in IL-1β induction or processing in inflammatory cells stimulated in vitro with Candida albicans [29] , [36] , [37] . IL-18 is another member of the IL-1 family which requires proteolytic enzyme cleavage for activation [57] , [58] . Although caspase-1 mediated cleavage of pro-IL-18 into the bioactive form is the accepted paradigm , alternative mechanisms for cleavage have been proposed including PR3 , granzyme B and mast cell chymase [59] , [60] . In our studies , in vivo IL-18 induction by Candida infection showed a similar pattern in mucosal tissues as IL-1β . Next , we investigated the regulation of the cytokine IL-17 , which has been shown to play a role in anti-fungal immunity in humans and in animal models [41] , [42] , [44] , [45] . In response to infection with Candida albicans , IL-17A , IL-17F and IL-17RA were up-regulated in mucosal tissues . Interesting patterns of induction of the IL-17 family by Candida infection were observed; IL-17A was dependent on NLRC4 , NLRP3 and ASC whereas IL-17F was only dependent on NLRP3 with comparable induction seen in WT , NLRC4 and ASC deficient mice . The induction of IL-17RA was only dependent on NLRC4 . The inflammasome dependence of IL-17 family responses during microbial infection is not well understood , and the regulation of these genes is likely multi-factorial , including dependence on IL-1β as well as other inflammatory mediators . The cytokine IL-6 was induced in wild-type mice in response to Candida infection but was significantly reduced in NLRC4 deficient mice and completely abrogated in NLRP3 and ASC deficient mice . The chemokine KC ( CXCL1 ) , a murine homolog of human IL-8 , was also found to be highly dependent on IL-1β as NLRC4 , NLRP3 , and ASC deficient mice exhibited significant decreases in gene expression following fungal challenge . Although IL-6 and KC/IL-8 are not known to be major mediators of anti-fungal immunity , they play an important role in the cytokine network of innate cellular communication . IL-6 has been shown to be directly up-regulated in macrophages by Candida cell wall components [61] but can also be induced by secondary effect of other inflammatory responses . IL-6 signaling has also been shown to be important in the recruitment of neutrophils in response to Candida [38] , [62] . IL-6 production is regulated by a complex network of signaling pathways that include NF-κB as well as a newly described pathway mediated by the protein tyrosine phosphatase , Src homology domain 2-containing tyrosine phosphatase-1 ( SHP-1 ) leading to activation of Erk1/2–C/EBPβ [63] . This is particularly relevant to mucosal anti-fungal immunity as SHP-1 , also known as PTPN6 and PTP1C , has been shown to negatively regulate the effects of epidermal growth factor [64] , an important regulator of epithelial homeostasis , as well as affect tight junction formation in epithelium [65] . We propose , based on our histological analysis , that the defect in the induction of inflammatory mediators observed in mice deficient in the NLRP3/ASC inflammasome is most likely due to defective functioning of inflammatory leukocytes in the absence of these proteins . In contrast , the partial defect in inflammatory responses observed in mucosal tissues from infected NLRC4−/− mice is likely due to impaired infiltration of immune cells into infected tissue . Another key innate immune response during infection is the release of antimicrobial peptides designed to limit pathogen growth and survival . Antimicrobial peptides ( AMPs ) consist of a diverse group of small cationic peptides including the defensins , cationic and amphipathic peptides which have broad antimicrobial and chemotactic properties . Beta-defensins are primarily secreted by epithelial cells and play an important role in the microbial homeostasis of the skin , oral cavity , lung and gut . Human β-defensin ( hBD ) -1 is primarily expressed in the urinary and respiratory tracts [66] , [67] and although constitutively expressed , may be up-regulated by infection or inflammation . A defect in hBD-1 activity in the lung has been associated with cystic fibrosis [68] , [69] . Polymorphisms in the defensin-1 gene , defB1 , have been associated with low oral colonization with Candida albicans ( Jurevic 2003 ) , protection from HIV [70]–[72] , chronic obstructive pulmonary disease [73] and Crohn's disease [74] . The murine homolog of hBD-1 , murine β-defensin ( mBD ) -1 , is also expressed by epithelial surfaces , lung and kidney and has salt sensitive antimicrobial activity [75] , although its role in antifungal defense is unclear . Both hBD-2 and hBD-3 have known anti-Candida [46] , [76]–[78] as well as anti-HIV activity [79] , [80] . The role of mBD-2 , the murine homolog of hBD-2 , in oral mucosal health is unclear although its expression in the lung is highly inducible by LPS [81] . The murine ortholog of hBD-3 , mBD-14 , has inducible expression in the respiratory and intestinal tracts as well as in dendritic cells and shows anti-bacterial and chemotactic activity [82] . To better define the role of AMPs in anti-fungal defense , we examined AMP responses in oral mucosa after infection with Candida albicans . We show that mBD-1 appeared to be constitutively expressed in WT mice; however , gene expression was inhibited in NLRP3 and ASC deficient mice . We discovered that mBD-2 , -4 and -14 were highly dependent on inflammasome activation as both NLRC4 and NLRP3 as well as ASC deficient mice exhibited dramatically reduced expression levels compared to WT . Interestingly , mBD-3 responses were found to have little dependence on NLRC4 but were dependent on NLRP3 and ASC . Another class of AMPs , the cathelicidins , consisting of human LL-37 and murine CAMP ( also known as CRAMP ) , are known to have anti-Candida as well as chemotactic activity [83]–[87] . We observed that CAMP was highly up-regulated in WT mucosa in infected mice , but not in NLRC4 , NLRP3 or ASC deficient mice . A recent report found that IL-17A augmented vitamin D3-mediated CAMP production in keratinocytes during psoriasis [88] . In concurrence with these findings , we observed abrogated CAMP expression in NLRC4 and NLRP3 deficient mice , which also lacked IL-17A gene expression , following Candida challenge . In addition to its direct antimicrobial effects , CAMP has been identified as a modulator of the P2X7R which has a known role in ATP-induced IL-1β release [49] , [89] , [90] . From our studies , it can be inferred that initial production of IL-1β may induce IL-17A and CAMP production which can in turn positively regulate further production of IL-1β to create an inflammatory environment which limits fungal infection . This mechanism may serve to explain the strong in vivo phenotype observed in our model for NLRC4 and NLRP3 deficient mice , perhaps via a failure to engage this positive feedback loop resulting in an immune state that is prone to persistent infection . In addition to driving IL-1β and IL-18 responses , the NLRC4 inflammasome has been shown to induce a specialized form of programmed cell death , termed pyroptosis or pyronecrosis , characterized by the release of cytoplasmic contents , which include inflammatory mediators such as ATP and arachidonic acid metabolites , to the extracellular matrix . A defect in pyroptosis may partially account for the critical role for NLRC4 activation in our model of candidiasis and provides an opportunity for future research . Interestingly , our data shows that activation of the NLRC4 inflammasome is important in the stromal compartment , where its role is critical for in vivo anti-fungal host defense , but not in the hematopoietic compartment . Using bone marrow chimera mice we differentiated between inflammasome activity in hematopoietic derived cells such as infiltrating macrophages and neutrophils , and embryonic derived mucosal tissues in our model of oropharyngeal candidiasis . This approach demonstrated that NLRP3 and ASC activity in both hematopoietic and stromal compartments are important for protection against oral infection and dissemination . This is in agreement with our previously published findings that the NLRP3 inflammasome was the primary mediator of IL-1β cleavage in murine macrophages stimulated with Candida in vitro [6] . In contrast , in vivo infection of bone marrow chimera mice showed higher mucosal colonization in NLRC4 deficient recipient mice reconstituted with WT inflammatory cells than WT recipients reconstituted with Nlrc4-/- cells , which had similar levels of oral mucosal infection as WT controls . Overall , these studies utilizing chimera mice in our murine model of mucosal fungal infection implicate a novel , tissue-specific role for the NLRC4 inflammasome . Many key questions are raised by the findings in this paper . Known microbial activators of the NLRC4 inflammasome include Salmonella typhimurium [91] , Shigella flexneri [92] , [93] , Legionella pneumophila [94] and Pseudomonas aeruginosa [34] . Previous reports implicated the activation of NLRC4 by the release of flagellin through the Type-III secretion apparatus and by components of the basal rod proteins of the Type III secretion system itself [93] , [95] , [96] . Despite these findings , it still remains unclear the mechanism by which NLRC4 senses these activators . Given the homology between the known bacterial activators of NLRC4 , it is possible that these proteins may function as a direct receptor recognizing a conserved sequence or structural feature . Current models of NLRP3 activation indicate it does not act as a traditional receptor but rather as a nexus for different pathways invoked following cellular injury and/or infection , which may also be true for the NLRC4 inflammasome . Future studies will seek to elucidate the mechanism of NLRC4 recognition of its activators and also identify the molecule ( s ) in Candida that induce NLRC4 activation . We hypothesize that mucosal NLRC4 activation may occur as an early event in fungal infection , perhaps as a result of cellular damage or direct effect of infection , leading to the induction of innate responses such as anti-microbial peptides and cytokines that recruit inflammatory cells including neutrophils and macrophages that infiltrate the sites of infection . Candida induced activation of the NLRP3/ASC inflammasome then provides a critical amplification of the innate response leading to protection of the host from overwhelming mucosal and disseminated candidiasis .
The animals described in this study were housed in the AAALAC accredited facilities of the Case Western Reserve University School of Medicine . All animal use protocols have been approved by the Institutional Animal Care and Use Committee of Case Western Reserve University and adhere to national guidelines published in Guide for the Care and Use of Laboratory Animals , 8th Ed . , National Academies Press , 2001 . Candida albicans strains GDH2346 ( NCYC 1467 ) , a clinical strain originally isolated from a patient with denture stomatitis , or ATCC 90234 were utilized for in vitro and in vivo studies . Master plates were maintained on Sabouraud Dextrose ( SD ) agar . For OPC infection , yeast were grown for 12–16 h in SD broth , pelleted at 3000 rpm for 5 min and washed 2x with sterile 1X PBS . Yeast cells were manually counted using a hemocytometer and diluted to 5×107 cells/mL for live infection . Wild-type C57BL/6 mice were purchased from Jackson Laboratories . Nlrp3−/− , Nlrc4−/− and Asc−/− mice were generated by Millenium Pharmaceuticals . Animals were housed in filter-covered micro-isolator cages in ventilated racks . Infection and organ harvesting was performed as described previously [6] . Briefly , after pre-treatment with antibiotic containing water , the mice were anesthetized and light scratches made on the dorsum of the tongue following by the introduction of 5×106 C . albicans yeast . The scratches are superficial , limited to the outermost stratum corneum , and do not cause trauma or bleeding . After infection of 3 to 21 d , the mice were euthanized , organs harvested and homogenized and fungal burdens assessed by growth on SD agar . For gross clinical score assessment , visual inspection of fungal burdens on the tongues was performed under a dissection microscope . A score of 0 indicates the appearance of a normal tongue , with intact light reflection and no visible Candida , a score of 1 denotes isolated patches of fungus , a score of 2 when confluent patches of fungus are observed throughout the oral cavity , and a score of 3 indicates the presence of wide-spread fungal plaques and erosive mucosal lesions . For assessment of inflammatory gene induction , buccal tissue was isolated from infected mice and immediately placed into a RNA stabilization reagent ( RNAlater , Qiagen ) . After homogenization in lysis buffer for 1 . 5 min using a bead-beater homogenizer ( Retsch ) , total RNA was isolated using PrepEase RNA Spin Kit ( USB/Affymetrix ) followed by conversion to cDNA using SuperScript III Reverse Transcriptase ( Invitrogen ) . Whole blood was collected via retro-orbital bleeding into EDTA pre-coated tubes , followed by centrifugation and removal of serum . Serum was stored at −80°C until used . Quantitative real-time PCR was done as described [6] . Specific primer sequences are listed in Table S1 . Cytokines were measured in serum by ELISA ( R&D ) . NCBI gene accession numbers are as follows: Nlrc4:NM_001033367 . 3; Nlrp3: NM_145827 . 3; Asc: NM_023258 . 4; Il1b: NM_008361 . 3; Il1r1: NM_008362 . 2; Il1rn: NM_031167 . 5; Il17a: NM_010552 . 3; Il17f: NM_145856 . 2; Il17ra: NM_008359 . 2; Il18: NM_008360 . 1; Cxcl1: NM_008176 . 3; Il6: NM_031168 . 1; Defb1: NM_007843 . 3; Defb2:NM_010030 . 1; Defb3: NM_013756 . 2; Defb4: NM_019728 . 4; Defb14: NM_183026 . 2; Camp: NM_009921 . 2 . Intact tongues are removed at necropsy , immediately immersed in Tissue Freezing Medium ( EMS ) and flash frozen in liquid nitrogen . After cryo-sectioning , 5 µm sections were fixed with10% formalin for 2 min then stained with Periodic Acid Schiff and Hematoxylin ( PAS/H ) . For immunofluorescent staining , the sections were blocked with 5% normal goat serum/PBS , stained with rat monoclonal anti-neutrophil primary antibody ( NIMP-R14; specific for Ly-6G and Ly-6C ) and Alexa Fluor 488-conjugated goat anti-rat secondary antibody ( Invitrogen ) and mounted in Vectashield containing DAPI ( Vector Laboratories ) . For quantitative analysis , images were taken using a Leica DMI 6000B inverted microscope , and the number of neutrophils in each section was digitally quantified using the imaging program MetaMorph ( Molecular Devices ) . Briefly , the number of pixels in a region containing the dorsal epithelial portion of the tongue was counted . A threshold value was then assigned which corresponds to a minimum fluorescent value of a neutrophil . The number of pixels at or above this threshold was determined and a percentage of fluorescent pixels determined by dividing by total overall number of pixels . Lethally irradiated mice ( exposed to a Cesium-139 γ-radiation source for a total full body dose of 900 rads ) received 5×106 bone marrow cells from pooled donor mice via tail vein injection and allowed to recover for 4 weeks . Data were analyzed using commercial software ( GraphPad ) and Student's two-sample independent t tests or Mann Whitney U tests were used for comparative statistical analysis of qPCR , ELISA , and quantitative fungal load data . Comparison of survival curves was done using a mean Logrank test . P values are presented when statistical significance was observed ( significance set at P≤0 . 05 at a confidence interval of 95% ) . | In this manuscript we describe a new role for a group of molecules termed the “inflammasome” that process key immune response proteins including interleukin-1-β . In previous work , we and others have shown that the NLRP3 inflammasome is important in protecting from severe fungal infections . We now show that , in addition to the NLRP3 inflammasome , a different inflammasome containing NLRC4 is also important in protecting against infection with Candida albicans , and appears to be functioning in the mucosal lining of the mouth and intestines , rather than in immune cells . Our research explains a new mechanism of mucosal immunity to fungal infections and has broad implications for developing new treatments against fungal infections , which are a serious cause of illness and death , particularly in immunocompromised persons . Additionally , this research may also lead to new ways to identify those individuals who are at the highest risk for serious fungal infections . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"medicine",
"infectious",
"diseases",
"model",
"organisms",
"immunity",
"innate",
"immunity",
"immunology",
"biology",
"microbiology",
"yeast",
"and",
"fungal",
"models",
"candida",
"albicans"
]
| 2011 | A Novel Role for the NLRC4 Inflammasome in Mucosal Defenses against the Fungal Pathogen Candida albicans |
Threespine stickleback fish offer a powerful system to dissect the genetic basis of morphological evolution in nature . Marine sticklebacks have repeatedly invaded and adapted to numerous freshwater environments throughout the Northern hemisphere . In response to new diets in freshwater habitats , changes in craniofacial morphology , including heritable increases in tooth number , have evolved in derived freshwater populations . Using a combination of quantitative genetics and genome resequencing , here we fine-mapped a quantitative trait locus ( QTL ) regulating evolved tooth gain to a cluster of ten QTL-associated single nucleotide variants , all within intron four of Bone Morphogenetic Protein 6 ( Bmp6 ) . Transgenic reporter assays revealed this intronic region contains a tooth enhancer . We induced mutations in Bmp6 , revealing required roles for survival , growth , and tooth patterning . Transcriptional profiling of Bmp6 mutant dental tissues identified significant downregulation of a set of genes whose orthologs were previously shown to be expressed in quiescent mouse hair stem cells . Collectively these data support a model where mutations within a Bmp6 intronic tooth enhancer contribute to evolved tooth gain , and suggest that ancient shared genetic circuitry regulates the regeneration of diverse vertebrate epithelial appendages including mammalian hair and fish teeth .
Finding the genes and ultimately the mutations that drive the evolution of animal form remains an important goal in biology [1] . The cis-regulatory hypothesis proposes that cis-regulatory changes are the most frequent substrate for morphological evolution because these mutations are more likely to bypass the negative pleiotropy typically generated by coding mutations in developmental regulatory genes [2] . Although many studies in a variety of organisms have found cis-regulatory alleles underlying morphological evolution , less is known about why or how cis-regulatory alleles are used [3 , 4] . For example , for genes found to have cis-regulatory alleles associated with evolved differences , whether coding mutations generate negative pleiotropy and/or reduced fitness remains largely untested in many natural populations . Teeth are a classic model system for studying organ development and evolution in vertebrates [5 , 6] . During tooth development , epithelial and mesenchymal cells reciprocally signal to each other , integrating dynamic BMP , TGF-β , FGF , SHH , Notch , Activin , EDA , and Wnt signals to orchestrate the formation of a mature tooth [7 , 8] . Bone Morphogenetic Protein ( BMP ) signaling plays multiple critical roles during tooth development . During tooth initiation , epithelial Bmp4 inhibits expression of Pax9 and Pitx2 , developmental markers of the forming tooth placode [9 , 10] . These results suggest an inhibitory role of BMP signaling on tooth development . However , several lines of evidence support an activating role of BMPs on tooth development . For example , exogenous Bmp4 can rescue tooth development in Msx1 mutant mice and accelerate tooth development in cultured tooth mandibles , suggesting an activating role of BMP signaling [11 , 12] . Furthermore , mice with dental epithelial ablation of the BMP receptor , Bmpr1a , or transgenic for a construct overexpressing a BMP antagonist , Noggin , in dental epithelium have tooth arrest at the bud and placode stage , respectively [13 , 14] . Together , these results suggest that there are both activating and inhibitory roles of BMP signaling during tooth development . However , the roles of many BMP signaling components are not fully understood . Furthermore , the genetic pathways of early tooth pattern and initiation have been extensively studied and well characterized in mice . Because mice are monophyodont rodents that do not replace their teeth , considerably less is known about the developmental genetic basis of tooth replacement . Polyphyodont vertebrates ( e . g . sharks , teleosts , and reptiles ) that continuously replace their teeth offer an opportunity to study the genetic and developmental basis of tooth regeneration [6] . Threespine stickleback fish ( Gasterosteus aculeatus ) are an excellent model for understanding the molecular genetic basis of natural variation , including evolved differences in tooth number [15 , 16] . Sticklebacks have undergone a dramatic adaptive radiation in which ancestral marine sticklebacks have colonized freshwater lakes and streams throughout the Northern hemisphere [17] . Recent genetic studies have implicated cis-regulatory changes of developmental signaling molecules as underlying several aspects of stickleback morphological evolution [18–23] . Genome-wide searches for regions under selection during freshwater adaptation have found an enrichment in non-coding elements of the genome , further implicating cis-regulatory changes in underlying stickleback evolution [24] . Freshwater sticklebacks have evolved several morphological adaptations in their head skeleton , some likely due to the shift to feeding on larger prey in freshwater niches [25] . While many freshwater adaptations in sticklebacks involve skeletal loss , a constructive gain of pharyngeal tooth number is seen in freshwater benthic ( adapted to lake bottom ) and creek populations [19 , 26] . Pharyngeal teeth lie in the pharynx of fish and are serial and phylogenetic homologs of mammalian oral teeth [27] . Pharyngeal jaw patterning is an adaptive trait in fish that covaries with diet and ecological niche [28] . Many aspects of the developmental genetic circuitry regulating tooth development are conserved from mice to fish [29–31] . Thus , evolved tooth gain in sticklebacks provides a powerful opportunity to understand the evolutionary genetics of tooth development and replacement . Evolved tooth gain in benthic freshwater fish from Paxton Lake in British Columbia is accompanied by an increase in the size of the tooth field , a decrease in tooth spacing , and an increase in tooth replacement rate late in development [19 , 26] ( Table 1 , columns 1–3 ) . Previously we showed that this derived tooth pattern is partially explained by a large effect quantitative trait locus ( QTL ) on chromosome 21 that is associated with a late-acting cis-regulatory downregulation of Bmp6 expression from benthic alleles in dental tissue [19] ( Table 1 , column 4 ) . These results make Bmp6 an excellent candidate gene for underlying evolved tooth gain by regulating tooth patterning and replacement . As no coding changes were found between marine and benthic freshwater alleles of Bmp6 [19] , we sought to map candidate regulatory regions of Bmp6 associated with evolved tooth gain . Here , we use a combination of recombinant mapping , comparative genomics , genome editing , and transcriptional profiling to further dissect the molecular genetic basis of evolved tooth gain and the role of Bmp6 during tooth development in threespine sticklebacks .
We previously identified and fine-mapped a large effect tooth number QTL to a 2 . 56 Mb 1 . 5-LOD interval on stickleback chromosome 21 containing an excellent candidate gene , Bone Morphogenetic Protein 6 ( Bmp6 ) , along with 58 other predicted genes [19 , 32] . To further fine-map this QTL , we identified three chromosomes with marine-benthic recombination events within the 2 . 56 Mb fine-mapped interval ( Fig 1 ) . Fish with each of these recombinant chromosomes were crossed to fish heterozygous for marine and benthic alleles of chromosome 21 to generate large ( >100 fish each ) crosses to test these recombinant chromosomes for effects on tooth number ( Fig 1A , S1 Table ) . Recombinant chromosomes that increase tooth number compared to marine chromosomes suggest that the tooth controlling region of chromosome 21 lies within the benthic portion of the recombinant chromosome . We used a likelihood ratio test to determine whether each recombinant chromosome behaved more like a marine or benthic chromosome . Recombinant chromosomes one and three increased tooth number , each behaving like a benthic allele of chromosome 21 ( P value from likelihood ratio test = 3 . 0 x 10−4 for both ) ( Fig 1B ) . Recombinant chromosome two did not increase tooth number , behaving like a marine allele of chromosome 21 ( P = 1 x 10−3 from likelihood ratio test ) ( Fig 1B ) . Together , these recombinant crosses support a new smaller genetic interval , 884 kb in the stickleback reference genome assembly [24] , that contains 21 predicted genes including Bmp6 ( Fig 1C ) , reducing the physical size of the interval and number of genes by 65% and 64% , respectively . To estimate the frequency of the chromosome 21 high tooth number allele within the wild Paxton benthic population , we generated six marine by benthic F2 crosses testing eight wild-derived benthic chromosomes ( named B1-B8 , Fig 2 , S2 Table ) . These chromosomes had different genotypes at three microsatellite loci located 5’ , within , and 3’ of the chromosome 21 tooth QTL , suggesting they are molecularly distinct wild chromosomes ( S2 Table , see Methods ) . We found that seven of these eight benthic chromosomes had significant effects on tooth number with the same direction and similar magnitude of effect ( Fig 2 , S2 Table ) . The benthic chromosome tested in cross 6 ( B8 ) had no effects on tooth number ( Fig 2 , S2 Table ) . These results together suggest that the high tooth number allele on chromosome 21 is at high frequency in the Paxton benthic population , but at least one lower-frequency benthic allele is not associated with an increase in tooth number . We hypothesized that the Paxton benthic chromosome 21 alleles that increase tooth number ( B1-7 , Fig 2 ) share sequence variants that underlie evolved tooth gain that are not present on marine alleles or the benthic chromosome 21 allele without the tooth QTL ( B8 , Fig 2 ) . To test for QTL-associated variants , we resequenced the genomes of the four benthic grandparents from crosses 1–4 , two F2 fish homozygous for chromosomes B7 and B8 , and the three marine grandparents from crosses 2 , 5 , and 6 tested in Fig 2 ( S3 Table ) . We identified 372 sequence variants ( consisting of 323 SNPs , and 49 indels ) within the 884 kb fine-mapped genetic interval that were present on all the benthic chromosomes with a large effect QTL , but not present on marine chromosomes ( Fig 3A ) . We gave variants a QTL concordance score: the absolute value of the proportion of times a variant allele is found in the benthic fish with a chromosome 21 tooth QTL minus the proportion of times the same allele was found in fish without a tooth QTL . Only ten of these variants ( all SNPs ) were perfectly associated with the presence of the tooth QTL ( Fig 3 ) . Strikingly , all of these variants lie within a ~4 . 4 kb region of Bmp6 intron 4 ( Fig 3B , S4 Table ) . We previously showed that a cis-regulatory decrease in expression of Bmp6 is associated with the chromosome 21 tooth QTL in Paxton benthic fish , suggesting that changes to Bmp6 regulatory elements underlie the tooth QTL [19] . We hypothesized that the region of intron 4 containing tooth QTL specific variants is a tooth enhancer of Bmp6 ( Fig 3B ) . To test for enhancer function , we cloned a ~2 kb intron 4 genomic fragment from marine fish into a reporter construct ( S5 Table ) . Transgenic fish for this construct expressed GFP in the distal tips of developing pectoral and median fins at eight days post fertilization ( dpf ) , and pharyngeal and oral teeth at 10 dpf ( S1 Fig ) . These domains have been previously shown to be endogenous sites of Bmp6 expression in developing sticklebacks [19 , 33] . These results demonstrate that the fourth intron of Bmp6 contains an enhancer active in developing teeth and fins . To define the minimally sufficient enhancer , we subcloned the ~2 kb fragment into two smaller fragments of ~1 . 3 kb and 511 bp based on patterns of sequence conservation ( Fig 4A , S5 Table ) , and tested for enhancer function in marine stickleback fish . The 511 bp construct is highly conserved in fish and contains no QTL-specific variants . The 1 . 3 kb construct includes the 511 bp region and a less conserved region that contains 6 of the 10 QTL-specific variants . The ~800 bp included in the ~1 . 3 kb construct but not the 511 bp construct drove no consistent expression , and no convincing differences were observed either between the ~1 . 3 kb construct and the 511 bp construct , or marine and benthic versions of the ~1 . 3 kb construct at early embryonic and larval stages [n > 3 injection rounds each , n > 20 GFP+ lenses ( the internal control domain driven by the Hsp70l promoter ) for both early embryonic and early larval comparisons] . Both the larger 1 . 3 kb construct and the 511 bp construct drove expression in the distal edges of the median and pectoral fins at eight dpf ( Fig 4B ) . By 13 dpf , the 511 bp enhancer drove expression in mesenchymal cells in developing pharyngeal teeth , as well as expression in the tooth epithelium ( Fig 4C ) . In developing teeth , the GFP-positive mesenchymal domain extended from each tooth germ deep into the tooth plate ( Fig 4C ) . This tooth expression continued into late juvenile stages when the pharyngeal tooth number differences arise between marine and freshwater populations ( Fig 4D ) [19] . GFP expression was also detected in late juvenile oral teeth ( Fig 4E ) . These results demonstrate that the intron 4 tooth QTL-associated variants surround an enhancer sufficient to drive expression in developing fins and teeth . We previously identified a TGFβ-responsive 5’ enhancer of Bmp6 that also drives expression in developing teeth and fins in sticklebacks [33] . Because stickleback Bmp6 expression is spatially and temporally complex in developing teeth [19] , we hypothesized that the two regulatory elements may control distinct aspects of Bmp6 expression in teeth . To test this hypothesis , we compared GFP expression patterns in fish stably transgenic for reporter genes for the 190 bp 5’ tooth enhancer or the 511 bp intron 4 tooth enhancer ( Fig 4C , 4F and 4G ) . As previously described [33] , we found that the 5’ enhancer drives robust expression in developing tooth epithelium and adjacent tooth mesenchyme ( Fig 4G ) . We found that the intron 4 enhancer drove expression that appeared distinct from the 5’ enhancer at some stages of tooth development ( Fig 4C and 4G ) . The intronic enhancer drove expression in the mesenchymal cores of mature teeth similar to the expression driven by the 5’ enhancer . However , the intronic enhancer drove deeper mesenchymal expression around the base of the developing tooth compared to the 5’ enhancer ( Fig 4C and 4G ) . To directly compare the tooth expression domains driven by the two enhancers , we generated fish transgenic for both a 511 bp intron 4 enhancer mCherry reporter construct as well as a 190 bp 5’ tooth enhancer [33] GFP reporter construct ( Fig 4H ) . The tooth expression domains were partially overlapping between the two enhancers in developing teeth ( Fig 4I and 4J ) . As was seen comparing the stable lines , both enhancers drive similar mesenchymal expression at early stages of tooth development , but the 5’ enhancer and not the intron 4 enhancer drove strong epithelial expression at these stages ( Fig 4I and 4I’ ) . As tooth development progresses , the intron 4 enhancer drove expression at the base of the mineralized tooth , in mesenchymal cells that did not express the 5’ enhancer ( Fig 4J and 4J’ ) . These results suggest that Bmp6 expression in tooth epithelial and mesenchymal cells is driven by at least two enhancers that drive partially overlapping yet distinct expression patterns . To test whether Bmp6 is required for tooth patterning in sticklebacks , we used transcription activator-like effector nucleases ( TALENs ) to generate two predicted loss-of-function mutations in stickleback Bmp6 ( Fig 5A , S6 Table ) . We designed a TALEN pair to target the highly conserved second exon of Bmp6 , which is 5’ to the exons encoding the predicted secreted ligand . Thus early stop codons would be predicted to generate strong loss-of-function alleles . Injection of these TALEN RNAs into stickleback embryos efficiently induced mutations in the Bmp6 target sequence . To identify mutations in F0 injected fish and mutations transmitted through the germline in F1 fish , we PCR amplified the surrounding sequence around the target site , digested this amplicon with EcoRI , then gel extracted and sequenced the uncut band ( S2 Fig ) . We found that 24–57% of injected F0 stickleback embryos had detectable deletions , with up to 12% of these embryos appearing to have biallelic mutations ( S2 Fig ) . Consistent with previous studies using TALENs in fish , we identified a spectrum of insertions and deletions at the target site ( Fig 5B ) [34] . We generated two mutant alleles that we bred to test for phenotypes: ( 1 ) a 13 bp deletion , and ( 2 ) a 3 bp deletion plus 4 bp insertion ( Fig 5B bold ) . Both of these mutations are predicted to produce frameshifts and an early stop codon 5’ to the secreted BMP ligand and thus are both likely strong loss-of-function alleles ( S3 Fig ) . To test for tooth patterning phenotypes in Bmp6 mutants , we intercrossed fish that were heterozygous for the 13 bp deletion or the 3 bp deletion plus 4 bp insertion and raised developmental time courses . Homozygous mutants were underrepresented from expected ratios at later developmental stages , suggesting early juvenile lethality ( S7 Table ) . The surviving homozygous mutants tended to be slightly smaller ( S7 Table ) . Because of the late stage lethality , we continued the Bmp6 mutant time course with heterozygous backcrosses for late juvenile and adult stages . To test for required roles of Bmp6 in tooth patterning , we quantified ventral pharyngeal tooth number , tooth plate area ( size of tooth field ) , and inter-tooth spacing , three phenotypes controlled by the chromosome 21 tooth QTL [19] in the Bmp6 mutant time course ( Fig 5C–5F ) . At the early juvenile stage , homozygous mutants had a reduction of both tooth number and tooth plate area compared to wild-type or heterozygous fish ( Fig 5D and 5E; Table 1 , column 5 ) . Beginning in early juveniles , heterozygous fish had fewer ventral teeth and smaller tooth plate area , which were both significantly more reduced at later time points including adults ( Fig 5D and 5E; Table 1 , column 6 ) . There were no significant differences in inter-tooth spacing at any stage ( Fig 5F ) . These results show that Bmp6 is required for specifying tooth number and the size of the tooth field . To test whether fish transheterozygous for both the 13 bp deletion and the 3 bp deletion/4 bp insertion have tooth patterning phenotypes , we generated a transheterozygote cross using the two different Bmp6 mutant alleles . We found that fish transheterozygous for the two different mutations had similar tooth patterning phenotypes as fish homozygous for the 13 bp deletion ( S8 Table ) . In addition to the bilateral ventral tooth plates , stickleback pharyngeal teeth are also present on two bilateral dorsal tooth plates , dorsal tooth plate 1 ( DTP1 ) and dorsal tooth plate 2 ( DTP2 ) [35] . We next asked whether Bmp6 also regulates dorsal pharyngeal tooth number . We found no significant differences in tooth number of either dorsal tooth plate at early developmental stages ( S4 Fig ) . In adults , DTP2 tooth number was significantly lower in heterozygous mutants , but to a lesser degree than the ventral tooth number differences at the same stage ( S4 Fig ) . For both dorsal tooth plates , tooth numbers trended in the same direction as seen for the ventral tooth plates , with fewer teeth in mutants than wild types . These results demonstrate that , like the chromosome 21 tooth QTL [32] , Bmp6 dosage has stronger effects on ventral pharyngeal tooth number than dorsal pharyngeal tooth number . To begin to identify the genetic networks downstream of Bmp6 , we performed RNA-seq of early juvenile wild-type and 13 bp deletion homozygous mutant bilateral pharyngeal tooth plates ( n = 3 of each genotype , S9 Table ) . Following read mapping and gene expression quantification , we performed principal component analysis of normalized read count of the entire dataset ( Fig 6A ) . PC1 explains a large fraction of the total variance ( 31 . 15% ) , and discriminates between the Bmp6 homozygous wild-type and mutant samples ( Fig 6A ) . Furthermore , genes whose expression correlated with the first principal component were highly enriched for gene ontology terms related to development and cell signaling ( S10 Table ) . To test whether stickleback Bmp6 regulates BMP target genes found in other systems , we compared the genes that were differentially expressed between wild-types and mutants to three different data sets , two from ToothCODE [8] and the third from a microarray study [36] . By combining literature mining of published mouse tooth development studies as well as their own functional analyses , the ToothCODE project collected a list of target genes downstream of BMP signaling in developing tooth epithelium or mesenchyme [8] . We tested whether stickleback orthologs of these epithelial and mesenchymal BMP target gene sets were differentially affected in Bmp6 mutant tooth plate tissue . Orthologs of mesenchymal BMP target genes as a whole displayed significantly reduced expression in Bmp6 mutants ( P = 1 . 25 x 10−2 ) , while orthologs of epithelial BMP target genes were not significantly affected ( Fig 6B ) . A third set of BMP signaling target genes was identified in a meta-analysis of published microarray studies [36] . We next asked whether stickleback orthologs of this gene set were significantly downregulated in Bmp6 mutant tooth plate tissue . We found this set of orthologs was significantly downregulated in Bmp6 mutants ( P = 3 . 12 x 10−4 ) , with 15/17 displaying a lower mean expression ( Fig 6B ) . These results show that stickleback Bmp6 is required to regulate a conserved battery of BMP-responsive genes . We hypothesized that the Bmp6 tooth number phenotype may result from changes in major signaling pathways known to be involved in tooth development [6 , 7] . The ToothCODE project manually curated a list of genes involved in tooth development in eight major signaling pathways ( BMP , FGF , SHH , Wnt , Activin , TGF-β , Notch , and EDA ) important for tooth development in mice [8] . We asked whether stickleback genes annotated as being in each of these pathways were concertedly differentially expressed in Bmp6 mutants compared to wild types . We found the TGF-β signaling pathway to be significantly downregulated ( P = 4 . 7 x 10−3 ) in Bmp6 mutant tooth plates ( Fig 6C ) . Strikingly , all eight TGF-β components tested had reduced mean expression in Bmp6 mutant tooth plates ( Fig 6D ) . In contrast , none of the other seven signaling pathways had significant expression differences ( Fig 6C ) , despite the differences in tooth number in Bmp6 mutants . Together these data suggest that Bmp6 positively regulates TGF-β signaling in stickleback tooth plate tissue . In polyphyodont sharks , fish , reptiles , and mammals , Sox2 has been implicated in putative epithelial stem cells during tooth replacement [37–39] . We found no significant differences in Sox2 expression between Bmp6 wild-type and mutant fish [mean FPKMs ( Fragments Per Kilobase of transcript per Million mapped reads ) of 91 and 97 , respectively] . In mice , Bmp6 inhibits the proliferation of hair follicle stem cells [40 , 41] . Teeth and hair are epithelial appendages with deep developmental and genetic homology [42–45] . Thus , we hypothesized that Bmp6 may play a conserved role of mediating stem cell quiescence during tooth replacement . A previous study characterized a set of hair follicle stem cell signature genes that are upregulated in the stem cell niche in the mouse hair follicle relative to the proliferating hair germ [46] . Bmp6 mutants showed a highly significant ( P = 8 . 5 x 10−12 ) decrease in the expression of stickleback orthologs of these genes ( Fig 6E and 6F ) . The reduced expression of the orthologs of these hair follicle stem cell signature genes supports the hypothesis that Bmp6 regulates stem cell quiescence during tooth replacement .
We previously identified a cis-regulatory downregulation of Bmp6 associated with the chromosome 21 tooth QTL [19] . Because there are no reported coding changes between marine and freshwater benthic alleles of Bmp6 in wild sticklebacks [19] , regulatory changes that change the spatiotemporal pattern and/or the quantitative levels of Bmp6 expression likely modulate natural variation in tooth patterning . Here we combined recombinant mapping and comparative genomics of multiple QTL crosses to fine-map this chromosome 21 tooth QTL to a haplotype within the fourth intron of Bmp6 . The association of ten variants in intron 4 of Bmp6 with the chromosome 21 tooth QTL , together with our data showing intron 4 contains a robust tooth enhancer , suggest a model in which these QTL-associated variants at least partially underlie the tooth QTL . Although all of the tooth QTL-associated mutations are outside of the minimally sufficient 511 bp tooth enhancer , we propose that some or all of these variants underlie the cis-regulatory changes in Bmp6 . One of the most outstanding questions for future research to address is whether these ten variants affect the spatiotemporal patterns and/or quantitative levels of enhancer activity ( Table 1 , column 7 ) . Although comparing the ~1 . 3 kb marine and benthic constructs has revealed no obvious differences at early embryonic and larval stages to date , several technical challenges including mosaicism in F0s and position effects in stable lines make comparing two enhancers in different fish difficult . Due to the dynamic and complex expression patterns of the intronic enhancer , addressing potential marine/benthic enhancer differences would be facilitated by better tools to precisely compare enhancer activity , either at the same integration site using transgene landing pads , or in the same fish using bicistronic constructs separated by an insulator . We note that the minimally sufficient 511 tooth enhancer contains a predicted Foxc1 binding site [47] . In mice , Foxc1 regulates mammalian hair regeneration in part through regulating BMP signaling and appears to directly regulate Bmp6 [48] , so potential FoxC inputs into Bmp6 expression in replacement teeth are especially intriguing . Of the marine/freshwater differences in the enhancer , one SNP alters a predicted NFATc1 binding site , a critical regulator of stem cell quiescence in the mouse hair follicle stem cell niche [49] . Another SNP affects a predicted Gli binding site , of interest because Gli expression is seen in multiple epithelial appendage stem cell niches in mice [50] . Future experiments will dissect what signals regulate this intronic enhancer , as well as what phenotypic consequences , if any , result from mutations in this enhancer . This intronic enhancer , like the 5’ tooth enhancer [33] , also drives embryonic and larval expression in developing pectoral and median fins . One interesting hypothesis these expression domains in fins raise is whether evolved differences in median or pectoral fin morphology are also regulated by this derived intronic haplotype . Perhaps supporting this hypothesis , a QTL regulating median fin morphology ( dorsal spine 3 length ) was previously mapped to a broad region of chromosome 21 overlapping Bmp6 [32] . Future experiments will also test whether the marine and freshwater versions of the enhancer have different activity in fins . Our transgenic assays show that the intronic enhancer of Bmp6 drives both overlapping and distinct domains of expression as the previously characterized 5’ Bmp6 enhancer [33] . Both enhancers drive overlapping expression in the mesenchymal cores of developing teeth . However , relative to the 5’ enhancer , the intronic enhancer also appears to drive deeper and broader mesenchymal expression and more restricted epithelial expression . These differences in expression patterns from the two enhancers suggest different signaling inputs control the mesenchymal and epithelial expression of Bmp6 in developing teeth . Our finding that the 5’ and intronic Bmp6 enhancers drive partially non-overlapping expression patterns is reminiscent of the mouse Bmp5 gene , which has two rib enhancers that drive expression in largely complementary patterns [51] . A modular cis-regulatory architecture is likely a common feature of Bmp genes , and could predispose these genes to frequently be used in morphological evolution [21 , 52–54] . This QTL confers late-acting ( juvenile stage , >25mm in fish length ) increases in tooth number and tooth field size , and decreases in tooth spacing [19] . Here we generated fish with induced mutations in Bmp6 to directly test whether Bmp6 played any required role in regulating tooth patterning . Strikingly , fish heterozygous for induced mutations in Bmp6 also had developmentally late differences in tooth number and tooth field size , similar to the tooth QTL ( Table 1 ) . A second phenotypic similarity between the tooth QTL and induced mutations in Bmp6 is a stronger effect on ventral tooth number than dorsal tooth number [19 , 32] . However , the direction of the cis-regulatory allele , where the high-toothed allele drives reduced Bmp6 expression in cis relative to a marine allele [19] , would predict that a mutation that lowers Bmp6 mRNA levels would increase tooth number , while the Bmp6 coding mutants have fewer teeth ( Table 1 ) . One explanation for this unexpected direction of effect could be a threshold effect: the Bmp6 mutations were made in a freshwater benthic genetic background with already reduced levels of Bmp6 expression , and further lowering of Bmp6 activity could inhibit tooth development . One test of this model could be to analyze the role of Bmp6 during tooth development in marine sticklebacks , or in other freshwater populations lacking the benthic Bmp6 intronic haplotype reported here . Alternatively , the induced mutant coding alleles of Bmp6 might not recapitulate the evolved cis-regulatory differences between marine and freshwater fish . The dynamic expression of Bmp6 in dental epithelium and mesenchyme at different stages of tooth development is controlled by at least two different cis-regulatory elements ( [33]; this study ) , which we show here drive expression at some stages in non-overlapping patterns . The evolved cis-regulatory allele of Bmp6 may change the spatiotemporal pattern and/or levels of Bmp6 mRNA in different tissues , leading to different phenotypes than the coding mutations . Inducing loss-of-function mutations in the two known stickleback Bmp6 enhancers and assessing potential changes in tooth patterning could test this hypothesis . The cis-regulatory hypothesis proposes that morphological evolution typically proceeds through cis-regulatory mutations that avoid the negative pleiotropy typical of coding mutations [1 , 2 , 55] . Recent studies have shown that cis-regulatory and coding mutations can drive morphological evolution , and that the type of mutation may depend on the degree of pleiotropy of the gene of interest [18 , 19 , 56 , 57] . The lethality and smaller size of fish homozygous for Bmp6 coding mutations could explain why cis-regulatory changes of Bmp6 have been used to evolve increases in tooth number . There were no significant differences in tooth pattern at early developmental stages between wild-type and heterozygous Bmp6 mutant fish . However , as these heterozygous fish continued to develop to adult stages , when newly forming teeth are likely replacement teeth , the reduction of tooth number and tooth plate area became more dramatic , suggesting that tooth development at late stages is more sensitive to the dosage of Bmp6 . These differences could be due to different developmental or genetic constraints at the early juvenile and late adult stages of tooth patterning . For example , there could be more functional redundancy of Bmp6 with other BMP ligands in teeth at early developmental stages that compensate in Bmp6 heterozygous mutants . Alternatively , these differences may signify differing roles of Bmp6 in primary and replacement tooth formation: later developing replacement teeth may be more sensitive to Bmp6 dosage than primary teeth . However , homozygous mutants had significantly fewer teeth at early juvenile stages , suggesting Bmp6 is also required for formation of primary teeth . To test which genes and pathways are downstream of Bmp6 signaling , we used RNA-seq to compare genome-wide transcriptional profiles of wild-type and homozygous mutant Bmp6 tooth plates . Seven signaling pathways were not significantly different in this contrast , perhaps surprising given the predicted difference in total tooth number in these samples . However , we found that there is a concerted downregulation of the TGF-β signaling pathway components in homozygous mutants . TGF-β signaling is required for tooth development [58–60] . Furthermore , TGF-β signaling regulates Bmp6 expression in stickleback teeth through the previously described 5’ tooth enhancer [33] . These results suggest that TGF-β signaling is involved both upstream and downstream of Bmp6 during tooth development . During tooth development in mice , reciprocal signaling events involving Bmp4 and Msx1 occur between developing tooth epithelium and mesenchyme: Bmp4 expression is first detected in dental epithelium , is required to induce Msx1 expression in underlying mesenchyme , which in turn is required to induce Bmp4 expression in dental mesenchyme [11 , 61–63] . Thus , Bmp4 is thought to play critical roles during tooth development in both dental epithelium and mesenchyme . A large mouse gene expression study revealed sets of genes regulated by Bmp2/4/7 in dental epithelium and mesenchyme [8] . We hypothesized that mouse BMPs and stickleback Bmp6 regulate a conserved set of downstream genes in developing teeth . We tested this hypothesis by asking whether orthologs of known mouse BMP signaling target genes are differentially regulated in stickleback Bmp6 mutant tooth plate tissue . Surprisingly , we found significantly reduced expression of the set of genes responsive to BMP signaling in mouse dental mesenchymal cells , while the set of genes responsive to BMP signaling in mouse dental epithelial cells was not significantly altered . Perhaps consistent with a relatively lesser effect on dental epithelia than mesenchyme in the Bmp6 mutant , Sox2 , implicated in epithelial stem cells during tooth replacement in other polyphyodonts [37–39] , was not significantly affected in Bmp6 mutants . In other vertebrates that undergo tooth replacement , dental stem cells have been proposed to mediate tooth replacement [37–39 , 64–66] . Teeth develop from placodes , transient epithelial thickenings that grow outwards or inwards to form epithelial appendages [42 , 43] . Teeth are developmentally deeply homologous to other placode-derived organs , such as mammalian hair [44 , 45 , 67] . Mammalian hairs , like fish teeth , are constantly replaced throughout adult life . During mammalian hair regeneration , Bmp6 regulates stem cell quiescence in the hair follicle stem cell niche [40 , 46] . Additionally , conditional knockout of the BMP receptor Bmpr1a in mouse hair follicles resulted in a loss of both hair regeneration and stem cell signature genes [46] . Thus , we hypothesized that stickleback Bmp6 might regulate similar genetic pathways during tooth replacement as during hair regeneration . Supporting this hypothesis , in Bmp6 mutant tooth plate tissue , we found a significant downregulation of mouse hair follicle stem cell signature genes , a set of genes previously described to be upregulated in mouse hair follicle stem cells compared to cells in the forming hair germ [46] . This result supports a model where modulating Bmp6 expression in derived freshwater sticklebacks alters dental stem cell dynamics to result in the elevated tooth replacement rate seen in high-toothed freshwater sticklebacks [26] . Furthermore , this result suggests that the genetic circuitry regulating stem cell quiescence in continuously regenerating mammalian hair may be shared during constant tooth replacement in fish . This shared gene set might reflect an ancient highly conserved pathway regulating vertebrate epithelial appendage regeneration . If so , further identifying this core conserved gene regulatory network would provide profound insights into vertebrate development , regeneration , and evolution .
All animal experiments ( including euthanasia by immersion in a buffered 250 mg/L tricaine methane sulfonate solution ) were done with the approval of the Institutional Animal Care and Use Committee from University of California , Berkeley ( protocol #R330 ) . Stickleback fish were raised in 29-gallon tanks in ~1/10th ocean water ( 3 . 5 g/l Instant Ocean salt , 0 . 4 mL/l NaHCO3 ) and fed live brine shrimp as larvae , then frozen daphnia , bloodworms , and Mysis shrimp as juveniles and adults . All fish crosses were conducted using artificial fertilization . Further F3-F5 generations of a Paxton Benthic freshwater by Little Campbell marine F2 cross [68] were propagated by intercrossing fish heterozygous for marine and benthic alleles of chromosome 21 ( identified by heterozygosity at Stn487 and Stn489 ) . Recombinant fish in F4- F5 generation were identified using microsatellite markers Stn487 and Stn489 which flank the genetic interval surrounding Bmp6 . Caudal fin tissue was genotyped by first isolating DNA by incubating for 20’ at 94°C , then digesting with 2 . 5 μL of 20mg/ml proteinase K in lysis buffer ( 10mM Tris , pH 8 . 3; 50 mM KCL; 1 . 5 mM MgCl2; 0 . 3% Tween-20 0 . 3% NP-40 ) for an hour at 55°C followed by 20 minutes at 94°C . One μl of undiluted DNA was used directly in the genotyping PCR . Once recombinant fish were identified , recombinant breakpoints were further mapped using a combination of microsatellite markers and restriction fragment length polymorphisms ( RFLPs ) . Primer sequences for the left and right markers used to refine each recombinant chromosome used in this study are shown in S1 Table . Gene content was determined by hand annotating the Ensembl predicted gene list . Recombinant fish were crossed to F4-F5 fish heterozygous for marine and benthic chromosome 21 that were also derived from the same F2 grandparents . The recombination events in crosses 1–3 were between markers Stn488 and Stn489 ( cross 1 ) , or between markers Stn487 and Stn488 ( crosses 2 and 3 ) . Genotypes of chromosome 21 in these three crosses were scored as M ( marine ) , B ( benthic ) , or R ( recombinant ) based upon the two locus genotypes of Stn488/Stn489 ( cross 1 ) or Stn487/Stn488 ( crosses 2 and 3 ) . Recombinant crosses were raised to ~30 mm standard length . Fish were stained for bone with Alizarin Red , cleared , and pharyngeal teeth were quantified as previously described [19] . If tooth number was significantly correlated with standard fish length , sex , or family , we corrected for each using a linear model and used residuals from that regression for statistical analysis ( S1 Table ) . To test whether each recombinant chromosome contained the tooth number QTL , we performed a likelihood-ratio test comparing two models , one with the recombinant chromosome behaving as a benthic chromosome and one with the recombinant chromosome behaving as a marine chromosome . Lab-reared stocks of Paxton Benthic fish used for F2 crosses were generated by incrossing wild-derived fish from Paxton Benthic lake , British Columbia . Five benthic fish were crossed to marine fish and F1s subsequently incrossed to generate six F2 crosses . The specifics of marine populations used in each cross are presented in S2 Table . Three microsatellite markers spanning the chromosome 21 tooth QTL were genotyped: CM1440 ( primer sequences 5’ to 3’: AAATGTGCTCCTGGATGTGC and CTTTCTCCTTCTGCCAAACG ) , Stn489 , and Stn488; this set of genotypes was used to define molecularly distinct chromosome 21s . F2 crosses 5 and 6 shared a benthic grandparent . This marker analysis suggests that there are eight molecularly distinct chromosome 21s in the five benthic grandparents . To determine the effect of chromosome 21 on tooth number , the F2 crosses were genotyped using microsatellites markers on chromosome 21 near the tooth QTL ( see S2 Table for details ) . The effects of fish size on tooth number were removed by linear regression and the residuals were back-transformed to the mean standard fish length in each cross . Statistical association between chromosome 21 genotype and back-transformed phenotypes was tested using an ANOVA in R . To determine if both benthic chromosomes had an effect on tooth number in each cross , we performed a likelihood-ratio test for each wild benthic chromosome comparing a model where that chromosome does not have an effect on tooth number to a model where both benthic chromosomes have an equal effect on tooth number . We resequenced the genomes of the four benthic grandparents from crosses 1–4 and F2 fish homozygous for chromosome B7 and B8 . We also sequenced the marine Little Campbell grandparents from crosses 5–6 , and the Japanese marine grandparent from cross 3 ( Fig 2 ) . Caudal fins were digested overnight at 55°C in Tail Digestion Buffer ( 10 mM Tris , pH 8 . 0 , 100 mM NaCl , 10 mM EDTA , pH 8 . 0 , 0 . 5% SDS , 10 μl of 20mg/ml proteinase K ) . Genomic DNA was purified with a phenol:choloroform extraction followed by ethanol precipitation . Genomic libraries were generated using the Nextera DNA Sample Prep Kit ( Epicentre Biotechologies ) , the Nextra DNA Sample Preparation Kit ( Illumina ) , or the Nextera XT DNA Library Preparation Kit ( Illumina ) . Paired-end reads ( 100 bp ) were sequenced using an Illumina HiSeq2000 . See S3 Table for details of library preparation and sequencing summary for each library . Resulting reads were aligned to the repeat masked verision of the reference stickleback genome [24] using the bwa aln and bwa sampe modules of the burrows-wheeler aligner [69] . As the genome assemblies in the minimal 884 kb meiotic interval are identical in the Jones et al . and Glazer et al . assemblies [24 , 70] , the original Jones et al . assembly was used [24] . Samtools ( version 0 . 1 . 17 ) [71] was used to create a sorted and indexed BAM file , and Picard tools ( version 1 . 51 ) ( http://broadinstitute . github . io/picard/ ) was used to fix mate information , add read groups , and remove PCR duplicates . GATK's Unified Genotyper ( parameters: '—genotype_likelihoods_model INDEL' , '-stand_call_conf 25' , and '-stand_emit_conf 25' ) RealignerTargetCreator , IndelRealigner ( parameter: '-LOD 0 . 4' ) was used to call potential target indels and perform realignment around indels . Base quality recalibration was accomplished using BaseRecalibrator . HaplotypeCaller ( parameters: '-emitRefConfidence GVCF' , '—variant_index_type LINEAR' , and '—variant_index_parameter 128000' ) was used to generating a genomic VCF ( gVCF ) file for each library . The resulting gVCFs were merged and variants were called using the GenotypeGVCFs module [72–74] . High quality variants were selected using the following criteria: 1 ) Variants must have a variant quality score greater than 400 . 2 ) Variants must not be called 'missing' or have a quality score of less than 10 in either high-coverage benthic genome . 3 ) Variants must not be called 'missing' or have a quality score of less than ten in no more than two genomes . To further remove stickleback specific repeats , we removed variants with >99% of the 100bp flanking sequence matching more than six places in the genome using blastn with an e-value of less than 1x10-30 [75] . QTL concordance score is the absolute value of the proportion of times a variant was present in benthic fish with a chromosome 21 tooth QTL minus the proportion of times the same variant was found in fish without a tooth QTL . QTL Concordance scores were calculated using a custom python script . To generate GFP reporter constructs , each of the intron 4 fragments from the Little Campbell marine grandparent from cross 5 was cloned upstream of the Hsp70l promoter in a Tol2 expression construct using NheI [33] . For the mCherry construct , we cloned mCherry into the Hsp70l reporter construct using SalI and ClaI and the inserts were cloned upstream using NheI and BamHI . Primers for construct generation and sequencing are shown in S5 Table . To generate transgenic stickleback , transposase messenger RNA was synthesized from pCS2-TP [76] plasmid linearized with NotI and transcribed using the mMessage SP6 in vitro transcription kit ( Ambion ) and purified using the Qiagen RNeasy column . One-cell marine stickleback embryos were injected with a mixture of 37 . 6 ng/μL plasmid DNA and 75 ng/μL RNA with 0 . 05% phenol red as previously described [33] . All transgene images presented are from stable lines except for the mCherry expression in Fig 4I and 4J and the ~2kb fragment in S1 Fig ( which were mosaic ) . To generate a TALEN pair to target the stickleback Bmp6 gene , we used the TAL effector Nucleotide Targeter 2 . 0 ( https://tale-nt . cac . cornell . edu/node/add/talen ) ) to scan the second exon sequence of Bmp6 for potential target sites [77 , 78] . We chose TALEN parameters as described [34] . We chose a target site that is unique to Bmp6 in the stickleback genome and contains a common restriction site , EcoRI , which can be used to detect molecular deletions . We assembled the two TALEN constructs using Golden Gate cloning into the destination vectors pCS2TALDD and pCS2TALRR and verified correct assembly using Sanger sequencing as described [34] . See S6 Table for the specifics of the Bmp6 TALEN design . 5’-capped mRNA for each TALEN pair was transcribed using the SP6 mMessage Machine ( Ambion ) after the TALEN plasmid templates had been linearized with NotI . Pooled TALEN mRNA was injected into one-cell PAXB freshwater benthic stickleback embryos at a concentration of 40 ng/μL for each mRNA with 0 . 05% phenol red . To genotype fish for TALEN induced mutations , DNA was extracted as described above from adult fish caudal fin tissue or homogenized whole 1–3 dpf embryos . Genotyping PCR was performed using forward primer 5’- ACAAGCCGCTAAAAAGGACA-3’ and reverse primer 5’- GCACGTGTGCATGCTTTAGA -3’ . The reaction profile for the NEB Phusion reaction was 98°C for 30 seconds , 39 cycles of 98°C for 10 seconds , 58°C for 15 seconds , 72°C for 30 seconds , followed by 72°C for 10 minutes . The PCR products were cut directly with EcoRI . The products from the wild-type and mutant alleles are cut and not cut , respectively , by this assay ( See S3 Fig ) . Dorsal and ventral pharyngeal tooth number was quantified on a DM2500 Leica microscope using a TX2 filter as previously described [19] . For both ventral and dorsal tooth counts , total tooth number equals the sum of the left and right sides ( of ventral and dorsal pharyngeal teeth , respectively ) . Tooth plate area and spacing of the ventral pharyngeal tooth plate were quantified from a gray scale image taken with a DFC340 FX camera on a Leica M165FC as previously described [19] . Area and spacing of the ventral pharyngeal tooth plates are the averages of the left and right tooth plate . Skeletal traits were binned by total fish length for three stages: early juvenile <27 mm , late juvenile 27–37 mm , and adults >37 mm . Ventral tooth plates from three wild-type and homozygous mutant ( for the 13 bp deletion allele ) Bmp6 female sticklebacks ( standard length ~25 mm ) were dissected , placed into TRI reagent ( Sigma-Aldrich ) on ice , ground with a disposable pestle , and frozen overnight at -80°C . The next day , RNA was extracted , isopropanol precipitated , and resuspended in DEPC-treated water . 200 ng of purified RNA was used with Illumina's Truseq Stranded mRNA Library Prep Kit to create sequencing libraries . The resulting bar-coded libraries were pooled and 100 bp paired end reads were generated using a single lane of an Illumina HiSeq2000 . Reads were mapped to the stickleback reference genome [24] using STAR ( parameters: '—alignIntronMax 200000' '—alignMatesGapMax 200000' '—outFilterMultimapNmax 8' ) [79] . BAM files were created , sorted , and indexed using Samtools ( version 0 . 1 . 17 ) [71] . Picard tools ( version 1 . 51 ) was used to fix mate information , add read groups , and remove PCR duplicates ( http://broadinstitute . github . io/picard/ ) . Using the Ensembl reference transcriptome [24] , transcripts were quantified using cuffquant version 2 . 2 . 1 ( parameters: '-u' '—library-type fr-firststrand' ) and normalized using cuffnorm [80 , 81] . Principal component analysis of the resulting transcript abundances was done using the PCA package of FactoMineR ( http://factominer . free . fr/index . html ) in R , and was plotted in R . GO term enrichment for genes ranked by expression correlation with the first principal component of the RNAseq expression matrix was performed using GOrilla [82 , 83] . Hierarchical clustering was done using Cluster3 . 0 ( parameters: '-l' '-cg a' '-g 2' '-e 0' '-m c' ) [84] , and the results were visualized using JavaTreeView ( version 1 . 1 . 6r4 ) [85] . Additional figures and analyses were done using custom python scripts and figures created using matplotlib . ToothCODE gene sets were downloaded from the ToothCODE database ( http://compbio . med . harvard . edu/ToothCODE/ ) . ToothCODE identified downstream targets of Bmp signaling by literature mining manipulations of Bmp2 , Bmp4 , and Bmp7 . Targets that were upregulated when BMP signaling increased or downregulated when BMP signaling was decreased were termed BMP target genes . Stickleback orthologs of mouse hair follicle stem cell signature genes , genes upregulated in the hair follicle bulge relative to the hair germ [46] were identified using Ensembl predictions . Statistical enrichment was done similar to the methods as previously described [86] . Each gene in a set was subject to a t-test , obtaining a list of z-scores . The null hypothesis , that the gene set displays no differential expression enrichment , ( i . e . t-test z-scores are drawn from a standard normal distribution ) was tested using a 1-sample t-test , with resulting P values subject to a Bonferroni correction . The significance cutoff for the 1-sample t-test was confirmed by creating a simulated null distribution , using 10 , 000 permutations of an equal number of genes as in each gene set , randomly chosen without replacement . Cutoff test statistic values were chosen by taking the values at the 100- ( 2 . 5/N ) and 2 . 5/N percentile in the simulated null distribution , where N was the number of hypotheses being tested . Analysis was done using a set of custom python scripts , available upon request . | Understanding how traits evolve in nature remains a fundamental goal in biology . Threespine stickleback fish offer a powerful system to address this question . Ancestral marine sticklebacks have colonized new freshwater environments , where new traits evolve , including increases in tooth number . This evolved increase in tooth number arises late in development and is associated with an accelerated tooth replacement rate in high-toothed freshwater fish . Using genetic and genomic data , here we mapped a genomic region regulating evolved tooth gain to an intronic region of Bone Morphogenetic Protein 6 ( Bmp6 ) . This intronic region contains a transcriptional enhancer that drives gene expression during tooth development and replacement . We induced mutations in a coding region of Bmp6 and found required roles for Bmp6 for viability , growth , and tooth patterning . By comparing genome-wide gene expression data in wild-type and Bmp6 mutant dental tissues , we found significant downregulation of a set of genes whose orthologs were previously shown to be expressed in quiescent mouse hair stem cells . Collectively these data support a model where mutations within a Bmp6 intronic tooth enhancer contribute to evolved tooth gain , and suggest that different vertebrate epithelial appendages such as teeth and hair regenerate using an ancient shared genetic program . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
]
| [
"medicine",
"and",
"health",
"sciences",
"fish",
"bmp",
"signaling",
"gene",
"regulation",
"vertebrates",
"marine",
"biology",
"marine",
"fish",
"animals",
"osteichthyes",
"freshwater",
"fish",
"digestive",
"system",
"genome",
"complexity",
"dentition",
"genomics",
"marine",
"and",
"aquatic",
"sciences",
"gene",
"expression",
"head",
"sticklebacks",
"signal",
"transduction",
"teeth",
"eukaryota",
"anatomy",
"cell",
"biology",
"physiology",
"jaw",
"earth",
"sciences",
"genetics",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"cell",
"signaling",
"digestive",
"physiology",
"introns",
"organisms"
]
| 2018 | An intronic enhancer of Bmp6 underlies evolved tooth gain in sticklebacks |
Recent advances in next-generation sequencing approaches have revolutionized our understanding of transcriptional expression in diverse systems . However , measurements of transcription do not necessarily reflect gene translation , the process of ultimate importance in understanding cellular function . To circumvent this limitation , biochemical tagging of ribosome subunits to isolate ribosome-associated mRNA has been developed . However , this approach , called TRAP , lacks quantitative resolution compared to a superior technology , ribosome profiling . Here , we report the development of an optimized ribosome profiling approach in Drosophila . We first demonstrate successful ribosome profiling from a specific tissue , larval muscle , with enhanced resolution compared to conventional TRAP approaches . We next validate the ability of this technology to define genome-wide translational regulation . This technology is leveraged to test the relative contributions of transcriptional and translational mechanisms in the postsynaptic muscle that orchestrate the retrograde control of presynaptic function at the neuromuscular junction . Surprisingly , we find no evidence that significant changes in the transcription or translation of specific genes are necessary to enable retrograde homeostatic signaling , implying that post-translational mechanisms ultimately gate instructive retrograde communication . Finally , we show that a global increase in translation induces adaptive responses in both transcription and translation of protein chaperones and degradation factors to promote cellular proteostasis . Together , this development and validation of tissue-specific ribosome profiling enables sensitive and specific analysis of translation in Drosophila .
Recent advances in next-generation sequencing such as RNA-seq have revolutionized the measurement and quantification of genome-wide changes in transcriptional expression , without pre-existing knowledge of gene identity , at unprecedented resolution [1 , 2] . In addition , biochemical tagging of ribosomes has emerged as a powerful way to provide insight into gene translation by separating the actively translating mRNA pool from overall mRNA abundance [3–8] , a technique termed TRAP ( Translating Ribosome Affinity Purification ) . Although this approach provides important insights into translational regulation , it lacks the resolution to differentiate between mRNA populations associated with few or high numbers of ribosomes , a distinction that can have major consequences for accurately defining translational rates [9] . This limitation was recently overcome through the development of a technique called “ribosome profiling” , which quantifies only mRNA fragments that are protected by ribosomes ( “ribosome footprints” ) . This enables the quantitative analysis of the number of ribosomes associated with each mRNA transcript , and is even capable of defining regions within RNA transcripts of ribosome association [10 , 11] . This technology has been used to reveal genome-wide adaptations to translation that would not have been apparent from transcriptional or translational profiling ( TRAP ) approaches alone [10 , 12–14] . However , despite the potential of ribosome profiling , this approach has not been developed for tissue-specific applications in Drosophila . The Drosophila neuromuscular junction ( NMJ ) is an attractive system to test the power of ribosome profiling . At this model synapse , sophisticated genetic approaches have revealed fundamental genes and mechanisms involved in synaptic growth , structure , function , and plasticity [15 , 16] . In particular , translational mechanisms contribute to synaptic growth , function , and plasticity at this synapse [17–19] . Indeed , a key role for translation has recently been implicated in mediating a form of synaptic plasticity intensively studied at the Drosophila NMJ referred to as Presynaptic Homeostatic Plasticity ( PHP ) . At this glutamatergic synapse , genetic loss of the postsynaptic receptor subunit GluRIIA leads to a reduction in the amplitude of miniature excitatory postsynaptic potentials ( mEPSPs; Fig 1A; [20] ) . However , the amplitude of evoked excitatory postsynaptic potentials ( EPSPs ) are maintained at wild-type levels due to an enhancement in the number of synaptic vesicles released ( quantal content ) . Thus , a retrograde signaling system is induced by loss of GluRIIA that ultimately potentiates presynaptic release , restoring baseline levels of synaptic transmission [21 , 22] . Recent forward genetic screening and candidate approaches have revealed the identity of several genes and effector mechanisms in the presynaptic neuron required for the homeostatic potentiation of presynaptic release [22–24] . However , very little is known about the postsynaptic signaling system that transduces a reduction in glutamate receptor function into a retrograde signal that instructs an adaptive increase in presynaptic release . Thus , ribosome profiling of the postsynaptic muscle may reveal the nature of the retrograde signaling system mediating PHP . There is emerging evidence to strongly suggest that translational mechanisms in the postsynaptic muscle play a key role in retrograde PHP signaling . In particular , pharmacologic or genetic inhibition of postsynaptic protein synthesis through the Target of Rapamycin ( Tor ) pathway and associated translational modulators disrupts the expression of PHP in GluRIIA mutants [25–27] . Interestingly , a constitutive increase in muscle protein synthesis through postsynaptic overexpression of Tor was also shown to be sufficient to trigger the retrograde enhancement of presynaptic release without any perturbation to glutamate receptors [25 , 27] . Further , ongoing and sustained postsynaptic protein synthesis is necessary to maintain PHP expression , as acute inhibition of protein synthesis in late larval stages is sufficient to block PHP expression in GluRIIA mutants [25 , 26] . While these results establish some of the first insights into the postsynaptic signal transduction system controlling retrograde PHP signaling , the putative translational targets involved , and to what extent transcriptional and/or post-translational mechanisms contribute to PHP signaling , remain unknown . Thus , ribosome profiling has the potential to illuminate the translational targets necessary for postsynaptic PHP signal transduction . We have developed and optimized a streamlined system that enables ribosome profiling from specific tissues in Drosophila . We first validate the success of this approach in defining translational regulation in the larval muscle , and reveal dynamics in translation that are distinct from overall transcriptional expression . Next , we highlight the superior sensitivity of ribosome profiling in reporting translational regulation over the conventional TRAP method . Finally , we utilize this ribosome profiling approach to assess translational changes during two cellular processes . First , we evaluate the contributions of transcriptional , translational , and post-translational mechanisms in the postsynaptic muscle that drive the retrograde signaling system underlying presynaptic homeostatic potentiation . Second , we distinguish adaptive changes in transcription and translation that are triggered following a chronic elevation in muscle protein synthesis . This effort has highlighted the unanticipated importance of post-translational mechanisms in ultimately driving retrograde PHP signaling and illuminated the dynamic interplay between modulations in gene transcription and translation as cells acclimate to elevated metabolic activity while maintaining cellular homeostasis .
To assess the postsynaptic retrograde signaling systems that drive presynaptic homeostatic potentiation ( PHP ) at the Drosophila NMJ , we focused on three genetic conditions ( Fig 1A ) . First is the wild-type control genotype ( w1118;BG57-Gal4/UAS-RpL3-3xflag ) , which serves as the control condition in which PHP is not induced or expressed . Second , null mutations in the postsynaptic glutamate receptor subunit GluRIIA ( GluRIIASP16;BG57-Gal4/UAS-RpL3-3xflag ) lead to a chronic reduction in mEPSP amplitude [20] . However , EPSP amplitudes are maintained at wild-type levels due to a homeostatic increase in presynaptic release ( quantal content ) following retrograde signaling from the muscle ( Fig 1A–1D ) . This serves as one condition in which we hypothesized that gene transcription , translation , and/or post-translational changes may have occurred in response to loss of GluRIIA , triggering the induction of retrograde signaling that drives PHP . Indeed , in GluRIIA mutants , genetic disruption of the translational regulator Target of rapamycin ( Tor ) blocks PHP expression , resulting in no change in quantal content and a concomitant reduction in EPSP amplitude [25] . Finally , postsynaptic overexpression of Tor in an otherwise wild-type muscle ( Tor-OE: UAS-Tor-myc/+;BG57-Gal4/UAS-RpL3-3xflag ) is sufficient to trigger PHP signaling , leading to increased presynaptic release and EPSP amplitude with no change in mEPSP or glutamate receptors ( Fig 1A–1D; [25] ) . Tor-OE therefore served as the final genotype in which PHP signaling was induced through Tor overexpression without any perturbation of postsynaptic glutamate receptors . We hypothesized that changes in translation , and perhaps even transcription , in GluRIIA mutants , which may also be apparent in Tor-OE , would illuminate the nature of the postsynaptic transduction system underlying homeostatic retrograde signaling at the Drosophila NMJ . To define genome-wide changes in mRNA transcription and translation in the postsynaptic muscle that may be necessary for PHP signaling , we sought to purify RNA from third instar larvae muscle in wild type , GluRIIA mutants , and Tor-OE ( Fig 1E ) . We then sought to define mRNA expression through three methods: Transcriptional profiling , translational profiling using translating ribosome affinity purification ( TRAP ) , and ribosome profiling ( Fig 1F ) . First , we approximated the muscle transcriptome using transcriptional profiling of total mRNA expression by isolating RNA from the dissected third instar body wall preparation . This is primarily composed of muscle tissue , but also contains non-muscle cells including epithelia ( Fig 1F ) . Following extraction of RNA from this preparation , we generated RNA-seq libraries using standard methods ( see Materials and Methods ) . Next , to define translational changes specifically in muscle cells , we engineered an affinity tag on a ribosome subunit under control of the upstream activating sequence ( UAS ) , which enables tissue-specific expression ( Fig 1E ) . This biochemically tagged ribosome could therefore be expressed in muscle to purify ribosomes , then processed to sequence only mRNA sequences associated with or protected by ribosomes ( Fig 1F ) . Affinity tagging of ribosomes enabled us to perform translational profiling ( TRAP: Translating Ribosome Affinity Purification ) , an established technique capable of detecting ribosome-associated mRNA [3 , 5 , 7–9] . Finally , we reasoned that this approach could be optimized to enable ribosome profiling , which has been used successfully to determine changes in translation efficiency , with superior sensitivity over TRAP , in a variety of systems [10 , 12 , 13 , 28] . However , ribosome profiling has not been developed for use in specific Drosophila tissues . Our next objective was to optimize a tissue-specific ribosome profiling approach . Ribosome profiling is a powerful approach for measuring genome-wide changes in mRNA translation . However , high quantities of starting material is necessary to obtain sufficient amounts of ribosome protected mRNA fragments for the subsequent processing steps involved [29] . Since this approach has not been developed for Drosophila tissues , we first engineered and optimized the processing steps necessary to enable highly efficient affinity purification of ribosomes and ribosome protected mRNA fragments by incorporating ribosome affinity purification into the ribosome profiling protocol . Although tissue-specific ribosome affinity purification strategies have been developed before in Drosophila [4 , 5 , 7] , these strategies have not been optimized to meet the unique demand necessary for ribosome profiling . Previous approaches tagged the same ribosome subunit ( RpL10A ) with GFP and 3xflag tags [4 , 5 , 7] , however , we found that these strategies lacked the efficiency necessary for ribosome profiling during pilot experiments . We thus set out to develop and optimize a new ribosome affinity purification strategy that enables the efficient purification and processing of ribosome-protected mRNA . First , we generated transgenic animals that express a core ribosome subunit in frame with a biochemical tag ( 3xflag ) under UAS control to enable expression of this transgene in specific Drosophila tissues ( Figs 1E and 2A ) . Therefore , based on high resolution crystal structures of eukaryotic ribosomes [30] , we selected alternative ribosomal proteins from the large and small subunits expected to have C terminals exposed on the ribosome surface . We cloned the Drosophila homologs of these subunits , RpL3 and RpS13 , in frame with a C-terminal 3xflag tag and inserted this sequence into the pACU2 vector for high expression under UAS control [31] . We then determined whether intact ribosomes could be isolated in muscle tissue following expression of the tagged ribosome subunit . We drove expression of UAS-RpL3-Flag or UAS-RpS13-Flag with a muscle-specific Gal4 driver ( BG57-Gal4 ) and performed anti-Flag immunoprecipitations ( Fig 2A ) . An array of specific bands were detected in a Commassie stained gel from the RpL3-Flag and RpS13-Flag immunoprecipitations , but no such bands were observed in lysates from wild type ( Fig 2B ) . Importantly , identical sized bands were observed in immunoprecipitates from both RpL3-Flag and RpS13-Flag animals , matching the expected distribution of ribosomal proteins [32] . The RPL3-Flag immunoprecipitation showed the same distribution as RpS13 but higher band intensity , indicating higher purification efficiency , so we used RpL3-Flag transgenic animals for the remaining experiments . In addition to ribosomal proteins , the other major constituent of intact ribosomes is ribosomal RNA . Significant amounts of ribosomal RNA were detected in an agarose gel from RpL3-Flag immunoprecipitates ( Fig 2C ) , providing additional independent evidence that this affinity purification strategy was efficient at purifying intact ribosomes . Next , we tested the ability of RpL3-Flag to functionally integrate into intact ribosomes . We generated an RpL3-Flag transgene under control of the endogenous promotor ( genomic-RpL3-Flag; S1A Fig ) . This transgene was able to rescue the lethality of homozygous RpL3 mutations ( S1A Fig ) , demonstrating that this tagged ribosomal subunit can integrate and function in intact endogenous ribosomes , effectively replacing the endogenous untagged RpL3 protein . Further , anti-Flag immunostaining of UAS-RpL3-Flag expressed in larval muscle showed a pattern consistent with expected ribosome distribution and localization ( S1B Fig ) . Next , we verified that muscle overexpression of RpL3-Flag did not lead to perturbations in viability , synaptic growth , structure , or function ( S1C–S1L Fig ) , nor did muscle overexpression of RpL3-Flag disrupt the expression of PHP in GluRIIA mutants or Tor-OE ( Fig 1A–1D ) . Finally , we confirmed that ubiquitous or muscle overexpression of RpL3-Flag did not induce phenotypes characteristic of flies with perturbed ribosome function such as the “minute” phenotype of inhibited growth ( [33]; S1A Fig ) . Thus , biochemical tagging of RpL3 does not disrupt its localization or ability to functionally integrate into endogenous ribosomes . Finally , we developed and optimized a method to process the isolated ribosomes and generate only ribosome protected mRNA fragments for RNA-seq analysis . First , we digested the tissue lysate with RNaseT1 , an enzyme that cuts single stranded RNA at G residues , together with anti-Flag affinity purification ( Fig 2D ) . Following digestion and purification , we ran RNA on a high percentage PAGE gel , excising the mRNA fragments protected from digestion by ribosome binding ( 30–45 nucleotides in length; Fig 2D ) . Sequencing of this pool of RNA demonstrated that the vast majority of reads mapped to the 5’UTR and coding regions of mRNA transcripts , with very few reads mapping to the 3’UTR of mRNA transcripts ( Fig 2E ) , where ribosomes are not expected to be associated . This coverage map also revealed heterogeneous distributions on mRNA transcripts with irregular and prominent peaks , as expected , which are indicative of ribosome pause sites on mRNA ( Fig 2E; [34] ) . In contrast , RNA-seq reads for transcriptional and translational profiling using TRAP mapped to the entire mRNA transcript with relatively even coverage ( Fig 2E ) . Extensive metagene analysis confirmed similar distributions around start and stop codons for genome-wide averaged RNA-seq reads ( S2 Fig ) . Importantly , replicate experiments demonstrated that this protocol generated highly reproducible measures of relative protein synthesis rates , defined by mRNA ribosome density , or the number of ribosome profiling Reads Per Kilobase of exon per Million mapped reads ( RPKM , also referred to as ribosome profiling expression value; Fig 2F ) . Thus , expression of RpL3-Flag enables the purification of ribosomes from specific tissues in Drosophila , and further processing reproducibly generates and quantifies ribosome protected mRNA fragments , which have been demonstrated to correlate with protein synthesis rates [11] . Translation can differ in significant ways from overall transcriptional expression through modulations in the degree of ribosome association with each mRNA transcript , in turn suppressing or enhancing protein synthesis rates [35] . Translation efficiency is a measure of these differences , defined as the ratio of translational to transcriptional expression [10] . Hence , translation efficiency ( TE ) reflects the enhancement or suppression of translation relative to transcriptional expression due to various translational control mechanisms [36] . Although both translational ( TRAP ) and ribosome profiling approaches can report TE , ribosome profiling should , in principle , exhibit superior sensitivity in revealing translational dynamics . We therefore compared translational and ribosome profiling directly to test this prediction . We compared TRAP and ribosome profiling to transcriptional profiling in wild-type muscle . In particular , we tested whether differences were apparent in the number of genes revealed to be translationally suppressed or enhanced through ribosome profiling compared to TRAP . We first analyzed the extent to which ribosome profiling and TRAP measurements correlate with transcriptional profiling by plotting the ribosome profiling and TRAP expression values as a function of transcriptional profiling ( Fig 3A and 3B; see Materials and Methods ) . A low correlation would indicate more translational regulation is detected , while a high correlation is indicative of less translational regulation . This analysis revealed a low correlation between ribosome profiling and transcriptional profiling ( correlation of determination r2 = 0 . 100; Fig 3A ) , while a relatively high correlation was observed between TRAP and transcriptional profiling ( r2 = 0 . 617; Fig 3B ) . Further , we subdivided all measured genes into three categories: high TE , medium TE , and low TE . These groups were based on translation efficiency as measured by ribosome profiling or TRAP , with high TE genes having a TE value >2 , low TE genes having a TE value <0 . 5 , and medium TE genes having a TE between 0 . 5 and 2 . This division revealed a higher number of genes in the high and low TE groups detected by ribosome profiling compared to TRAP ( Fig 3C ) . Together , these results are consistent with ribosome profiling detecting more genes under translational regulation compared to TRAP . We next investigated the genes under significant translational regulation ( genes with high TE or low TE ) , detected through either ribosome profiling or TRAP , to determine whether differences exist in the amplitude of translational regulation detected . Specifically , genes were divided into the three categories mentioned above based on the average translation efficiency measured by ribosome profiling and TRAP . We then determined the ration of the ribosome profiling TE to TRAP within the three categories . A ratio above 0 ( log2 transformed ) in the high TE group indicates a more sensitive reporting of translation for ribosomal profiling , while a ratio below 0 in the low TE group would also indicate superior sensitivity for the ribosomal profiling approach . This investigation revealed an average ratio of 0 . 28 within the high TE group , -0 . 15 within the medium TE group , and -1 . 25 within the low TE group ( Fig 3D ) . This analysis demonstrates that ribosome profiling is at least 22% more sensitive in detecting high TE , and 138% more sensitive in detecting low TE in comparison to TRAP . Thus , this characterization demonstrates that ribosome profiling provides a more sensitive and quantitative measurement of translational regulation in comparison to TRAP , validating this approach . Both subtle and dramatic differences have been observed in rates of mRNA translation relative to transcription , particularly during cellular responses to stress [12 , 37] . Having optimized and validated our approach , we went on to perform transcriptional and ribosome profiling in GluRIIA mutants and Tor-OE in addition to wild type ( S1 Table ) . To minimize genetic variation , the three genotypes were bred into an isogenic background , and three replicate experiments were performed for each genotype ( see Materials and Methods ) . We first determined the total number of genes expressed in Drosophila muscle , as assessed through both transcription and ribosome profiling . The fly genome is predicted to encode 15 , 583 genes ( NCBI genome release 5_48 ) . We found 6 , 835 genes to be expressed in wild-type larval muscle through transcriptional profiling , and a similar number ( 6 , 656 ) through ribosome profiling ( Fig 4A ) , with ~90% of transcripts being shared between the two lists ( S2 Table ) , indicating that the vast majority of transcribed genes are also translated . A subset of genes that appeared to be transcribed but not translated likely result from non-muscle RNA transcripts derived from the body preparation ( see Materials and Methods ) . Therefore , these transcripts were not analyzed further . We observed no significant differences in the size of the transcriptome and translatome between wild type , GluRIIA mutants , and Tor-OE . We then compared the muscle transcriptome to a published transcriptome from the central nervous system ( CNS ) of third-instar larvae [38] . This analysis revealed dramatic differences in gene expression between the two tissues ( Fig 4B ) . In particular , we found several genes known to be enriched in muscle , including myosin heavy chain , α actinin , and zasp52 , to be significantly transcribed and translated in muscle , as expected . In contrast , neural-specific genes such as the active zone scaffold bruchpilot , the post-mitotic neuronal transcription factor elav , and the microtubule associated protein tau , were highly expressed in the CNS but not detected in muscle ( S2 Table ) . Together , this demonstrates that the muscle transcriptome and translatome can be defined by the transcriptional and ribosome profiling strategy we developed with high fidelity . Next , we investigated genome wide translation efficiency distribution in larval muscle , and compared this with gene expression as assessed through transcriptional and ribosome profiling . We first calculated translation efficiency for all genes expressed in larval muscle and compared heat maps of TE to heat maps of the transcription and translation level ( Fig 4C ) . This revealed a dynamic range of translation efficiency , and a surprising trend of genes with high TE displaying relatively low transcriptional expression levels , while genes with low TE exhibited high transcriptional expression levels ( Fig 4C ) . We then analyzed the genes categorized as high TE , medium TE and low TE ( described above ) in more detail , comparing the relative distribution in transcriptional expression . We found this trend to be maintained , in that high TE genes exhibited significantly lower transcriptional expression , while low TE genes were significantly higher in transcriptional expression ( Fig 4D ) . Together , this implies a general inverse correlation between translational and transcriptional expression . Finally , we examined the genes with the most extreme enhancement or suppression of translation efficiency to gain insight into the functional classes of genes that exhibit strong translational control under basal conditions . Interestingly , amongst the genes with the most suppressed translation ( 100 genes with the lowest translation efficiency ) , we found a surprisingly high enrichment of genes encoding ribosome subunits and translation elongation factors ( Fig 4E and 4F; S3A Fig and S3 Table ) . Indeed , 73 of the 100 genes with the lowest translation efficiency were ribosome subunits , with all subunits exhibiting a consistently low TE , averaging 0 . 091 . Importantly , we confirmed that overexpression of RpL3-Flag does not change transcription of other ribosomal subunits , as quantitative PCR analysis of RpS6 transcript levels were not significantly different between wild type and RpL3-Flag overexpression animals ( 1 . 03±0 . 05 fold compared to wild type , n = 3 , p>0 . 05 , Student’s t test ) . In contrast , RpL3 , the subunit we overexpressed ( UAS-RpL3-Flag ) , was a clear outlier compared with the other ribosome subunits , showing a translation efficiency of 2 . 85 . This was expected due to the RpL3-Flag transcript containing artificial 5’ and 3’ UTRs optimized to promote high levels of protein synthesis [31] . This overall suppression in TE of ribosome subunits may enable a high dynamic regulatory range , enabling a rapid increase in production of ribosomal proteins under conditions of elevated protein synthesis . Consistent with this idea , we observed a coordinated upregulation of translation efficiency for ribosomal subunits when overall muscle translation is elevated in Tor-OE ( Fig 4H ) . This is in agreement with previous findings showing ribosome subunits and translation elongation factors as targets for translational regulation by Tor [39 , 40] . In contrast to the enrichment of ribosome subunits observed in the low TE group , diverse genes were found among the most translationally enhanced group , with genes involved in cellular structure being the most abundant ( Fig 4E and 4G; S3B Fig and S4 Table ) . These genes may encode proteins with high cellular demands , being translated at high efficiency . Indeed , counter to what was observed in genes with low TE , genes with high TE showed no significant change in TE following Tor-OE ( Fig 4H ) . Together , this analysis reveals that translation differs in dramatic ways from overall transcriptional expression , reflecting a highly dynamic translational landscape in the muscle . We confirmed the fidelity of our transcriptional and ribosome profiling approach by examining in molecular genetic detail the two manipulations we utilized to trigger postsynaptic retrograde signaling . The GluRIIASP16 mutation harbors a 9 kb deletion that removes the first half of the GluRIIA locus as well as the adjacent gene , oscillin ( Fig 5A; [20] ) . Analysis of both transcriptional and ribosome profiling of GluRIIASP16 mutants revealed no transcription or translation of the deleted region , as expected ( Fig 5A ) . Transcription and translation of the adjacent gene , oscillin , was also negligible ( wild type vs . GluRIIA: transcription = 15 . 9 vs . 0 . 08 RPKM; translation = 9 . 8 vs . 0 . 4 RPKM ) . However , the 3’ portion of the GluRIIA coding region was still transcriptionally expressed in GluRIIA mutants , while a significant reduction in translation was observed by ribosome profiling ( Fig 5A ) . Together , this confirms that although the residual 3’ region of the GluRIIA locus was transcribed , likely through an adjacent promoter , this transcript was not efficiently translated . Indeed , the peak ribosome profiling signals , which represent ribosome pause sites on the mRNA transcript , is known to be conserved for specific open reading frames [34] . However , this pattern was altered in GluRIIA mutants compared to wild type ( Fig 5A ) , suggesting the translation of the residual 3’ region of GluRIIA in GluRIIASP16 mutants was not in the same reading frame as the intact transcript . Thus , both transcriptional and ribosome profiling confirms that GluRIIA expression is abolished in GluRIIASP16 mutants . Next , we examined the expression of endogenous ( genomic ) and transgenically overexpressed ( UAS ) Tor through transcriptional and ribosome profiling . While both endogenous Tor and UAS-Tor mRNA share the same coding region , the 5’UTR and 3’UTR regions differ between these transcripts ( Fig 5B ) , enabling us to distinguish expression between these transcripts . We first confirmed a large increase in the expression of the Tor coding region in Tor-OE through both transcriptional profiling ( 68 fold ) and ribosome profiling ( ~1200 fold; Fig 5B , black ) . In contrast , expression of the endogenous 5’ and 3’ UTRs of Tor was similar between UAS-Tor and wild type ( Fig 5B , grey ) , while a dramatic increase in the expression of the UTRs of UAS-Tor was observed through both transcription ( 125 fold ) and translation ( 1200 fold; Fig 5B , red ) . Indeed , the translation efficiency of Tor was increased 14 fold in Tor-OE , consistent with the known influences of engineered 5’UTR and 3’UTR sequences in promoting translation in UAS constructs [41] . This analysis defines the levels at which Tor transcription and translation are enhanced when UAS-Tor is overexpressed in the Drosophila larval muscle , and further serve to validate the sensitivity of ribosome profiling . Finally , we sought to define whether Tor-OE induced a global elevation in translation and to determine whether a similar global shift may have also occurred in GluRIIA mutants . Indeed , most if not all mRNAs are capable of being translationally modulated by Tor , with Terminal OligoPyrimidine tract ( TOP ) mRNAs being the most sensitive to Tor regulation [40 , 42] . First , we confirmed a global shift in translation in Tor-OE compared to wild type , as expected given the role of Tor as a general regulator of Cap-dependent translation initiation [43] . We plotted a gene count histogram of Tor-OE versus wild type fold change measured by ribosome profiling , and overlaid the graph over a wild type over wild type ribosome profiling fold change histogram . A shift in global translation was observed in Tor-OE , with an average of 1 . 6 fold change in translation compared to 1 . 09 for wild type ( Fig 5C ) . This shift is significant when tested by Kolmogorov–Smirnov test ( p<0 . 001 ) ( Fig 5D ) . We then performed this same analysis for GluRIIA vs WT . However , we observed no significant shift in translation in GluRIIA ( 0 . 97 fold change compared to 1 . 09; Fig 5C ) . Thus , while Tor-OE induces a global increase in translation , loss of the GluRIIA receptor subunit in muscle does not measurably change overall translation . Given the substantial evidence that Tor-mediated control of new protein synthesis in the postsynaptic cell is necessary for retrograde PHP signaling [25] , we compared transcriptional and translational changes in muscle between wild type , GluRIIA mutants , and Tor-OE . We anticipated a relatively small number of transcriptional changes , if any , between these genotypes , while we hypothesized substantial differences in translation would be observed in both GluRIIA mutants and Tor-OE . The elevated translation of this exceptional subset of targets would , we anticipated , initiate postsynaptic PHP signaling and lead to an instructive signal that drives the retrograde enhancement in presynaptic release . Alternatively , we also considered the possibility that Tor-mediated protein synthesis may act in a non-specific manner , increasing overall protein synthesis in the postsynaptic cell , while there would be no overlap in translational changes between GluRIIA mutants and Tor-OE . In this case , post-translational mechanisms would operate on a global elevation in protein expression in Tor-OE , sculpting the proteome into an instructive retrograde signal . Indeed , the acute pharmacological induction and expression of PHP does not require new protein synthesis [44] , providing some support for this model . We therefore compared transcription and translation in wild type , GluRIIA mutants , and Tor-OE . We first compared transcription and translation in GluRIIA mutants and Tor-OE relative to wild type by plotting the measured expression values for each condition and determining the coefficient of determination , r2 . An r2 value equal to 1 indicates no difference between the two conditions , while a value of 0 implies all genes are differentially expressed . This analysis revealed a high degree of similarity between wild type and GluRIIA mutants in both transcription and translation , with r2 values above 0 . 98 ( Fig 6A , left ) . In contrast , a slightly larger difference exists in transcription between Tor-OE and wild type , with r2 = 0 . 920 ( Fig 6B , left ) . Although transcription should not be directly affected by Tor-OE , this implies that perhaps an adaptation in transcription was induced in the muscle in response to chronically elevated translation . Finally , translational differences were the largest between Tor-OE and wild type , with r2 values equal to 0 . 363 ( Fig 6B ) . Indeed , 2 , 352 genes showed changes greater than 1 . 5 fold in their measured RPKM compared to wild type in this condition ( S6 Table ) . This global analysis demonstrates there are very few transcriptional and translational changes in GluRIIA compared to wild type , while moderate transcriptional and robust translational changes exist in Tor-OE . Unexpectedly , in depth analysis of the transcriptome and translatome in GluRIIA muscle revealed that no genes were significantly altered . In particular , we eliminated genes that were up- or down-regulated due to known or expected influences in the genetic background ( GluRIIA and oscillin expression , and closely linked genes to this locus; see Materials and Methods ) . Using a standard cut off for expression , we found no gene to have a significant up-regulation in TE more than 2 fold in GluRIIA mutants ( Fig 6A , right ) . Even with a lowered threshold for significant expression changes ( >1 . 5 fold change ) , we observed only 5 genes transcriptionally upregulated and 1 gene translationally upregulated in GluRIIA versus WT ( Fig 6A , right . S5 Table ) . Given this small number at such a lowered threshold , we considered the possibility that the genetic background may influence expression of these genes . Consistent with this idea , all 6 upregulated genes are closely linked to the GluRIIA locus or were located on the X chromosome , areas we could not fully outcross to the isogenic line ( Materials and Methods and S5 Table ) . Although we cannot rule out transcripts with more subtle differences in translation ( below 1 . 5 fold ) or genes with very low and/or highly variable expression that may nonetheless contribute to translational regulation in GluRIIA mutants , the sensitivity of ribosome profiling enables us to conclude that no major changes in transcription or translation are present in the postsynaptic muscle of GluRIIA mutants . While no specific translational targets were identified to significantly change in GluRIIA mutants , we did identify 47 genes ( including Tor itself ) that exhibited significant increases in translation efficiency in Tor-OE ( >2 fold; Fig 6B , right and S6 Table ) . Among these 47 genes , 7 encode TOP RNAs [39] , including ribosome subunits ( Fig 6C ) . Tor-dependent translational control directly regulates TOP RNAs [39 , 40 , 42] , ribosome profiling was successful in identifying genes in this class . Given the striking finding that very few genes appear to be under transcriptional or translational control in the postsynaptic muscle of GluRIIA mutants , we considered the possibility that the 47 genes we identified to be translationally upregulated in Tor-OE may also show a parallel trend in GluRIIA mutants but below statistical significance . We therefore generated a heat map of these 47 genes in Tor-OE vs WT and compared this to the same 47 genes in GluRIIA vs WT ( Fig 6C ) . This analysis revealed no particular trend or correlation in GluRIIA among the 47 genes with increased translation efficiency in Tor-OE ( Fig 6C ) . Together , these results suggest that retrograde signaling in the postsynaptic muscle , induced through loss of GluRIIA , does not alter translation of a specific subset of targets , while Tor-OE induces a global , non-specific increase in translation . Thus , post-translational mechanisms are likely to confer the specific signaling processes that ultimately instruct retrograde PHP communication . Although Tor -OE is not expected to directly impact transcription , our analysis above indicated that transcriptional changes are induced following the global increase in translation by Tor-OE ( Fig 6B ) . This suggested that adaptations in transcription , and perhaps also translation , may have been triggered in Tor-OE in response to the cellular stress imparted by the chronic , global increase in muscle protein synthesis . Indeed , proteome homeostasis ( proteostasis ) is under exquisite control [45] , and sustained perturbations in Tor activity induces transcriptional programs that adaptively compensate to maintain proteostasis [46 , 47] . We therefore reasoned that by examining the changes in transcription and translation induced by Tor-OE , we may gain insight into how a cell adapts to the stress of chronically elevated translation . Transcriptional and ribosome profiling revealed 11 genes with significantly upregulated transcription ( fold change>3 and adjusted p-value<0 . 05; Fig 7A and S7 Table ) , and 75 genes with significantly upregulated translation ( fold change>3 and adjusted p-value<0 . 05; Fig 7A and S7 Table ) in Tor-OE compared to wild type . Interestingly , 8 of these genes exhibited shared increases in both transcription and translation ( Fig 7A ) , with their translational fold change ( revealed by ribosome profiling ) being larger than would be expected by their transcriptional fold change alone . This suggests a coordinated cellular signaling system that adaptively modulates both transcription and translation in response to the global elevation in translation following overexpression of Tor in the muscle . Further analysis revealed these upregulated genes to belong to diverse functional classes ( Fig 7B ) . Notably , we observed a striking enrichment in genes encoding heat shock proteins and chaperones ( GO term fold enrichment of 45 . 13 , p-value = 0 . 006; GO enrichment test; S4 Fig ) , factors known to assist with protein folding and to participate in the unfolded protein response , particularly during cellular stress [48 , 49] . Indeed , among the 7 heat shock protein genes with significant expression in the muscle ( S7 Table ) , 5 were significantly upregulated in translation and 3 were significantly upregulated in transcription , with the remaining 2 showing a strong trend towards upregulation ( Fig 7C and 7D and S7 Table ) . We performed quantitative PCR as an independent approach to verify the upregulation of heat shock proteins , which confirmed upregulation in the level of total mRNA and ribosome-associated mRNA ( S4B and S4C Fig ) . Given the well documented role for heat shock proteins in regulating protein folding , stability , and degradation in conjunction with the proteasome system [48] , this adaptation likely contributes to the stabilization of elevated cellular protein levels resulting from Tor-OE . Thus , the coordinated upregulation of heat shock proteins is one major adaptive response in transcription and translation following Tor-OE . In addition to heat shock proteins , we also identified genes involved in other cellular functions that are upregulated in Tor-OE and appear to enable adaptive responses to elevated cellular protein synthesis . For example , the E3 ubiquitin ligase subunit APC4 , involved in proteasome-dependent protein degradation [50] , was upregulated in Tor-OE ( Fig 7D ) . Interestingly , proteasome subunits were reported to be upregulated in cells with increased Tor activity [47] . We also identified the RNA polymerase subunit rpi1 and transcription factor myc to be upregulated following Tor-OE ( Fig 7D ) . These genes promote ribosome biogenesis , with RpI1 necessary to synthesize ribosomal RNA and Myc involved in promoting the expression of ribosome assembly factors [51 , 52] . Together , RpI1 and Myc likely promote the generation of additional ribosomes to meet the increased demands of protein synthesis induced by Tor-OE , consistent with previous studies showing Tor inhibition leads to decreased RpI1 transcription [53] . Hence , transcriptional and ribosome profiling defined adaptations in gene expression and protein synthesis that maintain proteostasis following chronic elevation in protein synthesis .
We have developed a highly efficient affinity tagging strategy and optimized RNA processing to enable tissue-specific ribosome profiling in Drosophila . Ribosome profiling has major advantages over measuring total mRNA expression and ribosome-associated mRNA ( translational profiling using TRAP ) . Profound differences can exist between transcriptional expression and actual protein synthesis of genes expressed in a tissue . RNA-seq of total mRNA ( transcriptional profiling ) does not capture translational dynamics [54 , 55] . Translational profiling using TRAP does provide insights into translation [9] , but is less sensitive in detecting translational dynamics compared to ribosome profiling , which accurately quantifies the number of ribosomes associated with mRNA transcripts ( Fig 3; [56] ) . One major obstacle that limited the development of tissue-specific ribosome profiling is the relatively large amount of starting material necessary to generate the library for next generation sequencing . Because only ~30 nucleotides of mRNA are protected from digestion by an individual ribosome [10] , ribosome profiling requires much more input material compared to standard RNA-seq [29] . Thus , the purification efficiency of the ribosome affinity tagging strategy and subsequent processing steps are very important to enabling successful profiling of ribosome protected mRNA fragments in Drosophila tissues . We achieved this high purification efficiency by systematically testing and optimizing multiple ribosome subunits ( RpL3 , RpL36 , RpS12 , RpS13 ) and affinity tags ( 6xHis , 1xFlag , 3xFlag ) , finally settling on the RpL3-3xflag combination to enable the highest purification efficiency ( Fig 2B and see Materials and Methods ) . Collectively , this effort differentiates our strategy from previous approaches in Drosophila that achieved ribosome profiling but lacked tissue specificity [12] or purified ribosome-associated RNA from specific tissues but lacked the ability to quantify ribosome association with mRNA transcripts [4 , 5 , 7] . This optimized ribosome profiling approach has illuminated genome-wide translational dynamics in Drosophila muscle tissue and demonstrated two opposing protein production strategies utilized in these cells: high transcriptional expression coupled with low translation efficiency , which was apparent for genes encoding ribosomal subunits ( Fig 4F ) , and low transcriptional expression coupled with high translation efficiency , which was observed for genes encoding proteins belonging to diverse functional classes ( Fig 4G ) . These complementary strategies are likely tailored towards different cellular needs , enabling modulatory control of nuclear gene transcription and cytosolic protein synthesis . Thus , transcriptional and ribosome profiling of muscle tissue has revealed that translational control of ribosomal protein synthesis may be a strategy tailored to the unique metabolic needs of this tissue . We have used transcriptional and translational profiling to determine the contributions of transcription and translation in the postsynaptic signaling system that drives the retrograde enhancement of presynaptic efficacy . Strong evidence has suggested that protein synthesis is modulated during homeostatic signaling at the Drosophila NMJ , with genetic disruption of Tor-mediated protein synthesis blocking expression and activation of the Tor pathway triggering expression [25–27] . We had expected that translational profiling would discover targets with increased translation efficiency in the muscles of GluRIIA mutants and/or following postsynaptic Tor overexpression , genetic conditions in which presynaptic homeostatic plasticity is chronically activated . However , no specific changes in transcription or translation were observed in GluRIIA mutants , while a large percentage of muscle genes increased in translation following Tor-OE ( Fig 5C and 5D ) . Furthermore , an apparent global increase in translation also appears sufficient to instruct enhanced presynaptic release , consistent with the nature of the translational regulators implicated in PHP: Tor , S6 Kinase , eIF4E , and LRRK2 [25 , 27] . These factors are cap-dependent translational regulators that act on nearly all mRNAs , although there is some degree of differential sensitivity of mRNAs to cap-dependent translational regulation [40 , 42] . Although we cannot rule out very subtle changes in translation , nor can we accurately measure levels of transcription or translation in genes with very low or highly variable expression , the sensitivity of the ribosome profiling approach rules out major changes in the translation of specific genes being necessary to promote PHP transduction . Thus , a global enhancement of translation may initiate post-translational mechanisms that are likely to ultimately drive PHP signaling in Tor-OE . Indeed , a recent study demonstrated that GluRIIA- and Tor-OE-mediated PHP ultimately converge at a post-translational mechanism to mediate the same retrograde signaling pathway [57] . We consider several possible explanations and implications of these findings . There are three conditions that trigger homeostatic retrograde signaling in the postsynaptic muscle: Acute pharmacological blockade of GluRIIA-containing postsynaptic receptors [44] , genetic mutations in GluRIIA [20] , and chronic overexpression of Tor [25] . First , all three manipulations lead to a similar enhancement in presynaptic release and converge to drive the same unitary retrograde signaling system [57] . Further , the acute pharmacological induction of PHP does not require new protein synthesis [44 , 57] . This implies that while distinct pathways mediate PHP signaling , they all ultimately converge on the same pathway that utilizes post-translational mechanisms . Indeed , there is evidence for post-translational mechanisms in the induction of PHP signaling in GluRIIA mutants , as changes in CamKII phosphorylation and activity have been observed [57 , 58] . In addition , other post-translational mechanisms , such as protein degradation or ubiquitination , could contribute to homeostatic signaling in the muscle . However , while all three manipulations appear to ultimately utilize the same retrograde signal transduction system , it is quite intriguing that somehow the global increase in translation observed in Tor-OE is sculpted , perhaps by shared post-translational mechanisms , into a specific retrograde signal that instructs enhanced presynaptic release . Second , it is possible that pharmacological , genetic , or Tor-OE-mediated inductions of PHP signaling are all mechanistically distinct , in which case no common transcriptional , translational , or post-translational mechanisms would be expected . Indeed , forward genetic screening approaches to discover genes necessary for PHP expression have failed to identify any genes needed for PHP induction in the postsynaptic muscle [23 , 59] , suggesting possible redundancy in these signaling systems . Further , it is possible that very small , local changes in translation are necessary to drive retrograde signaling in GluRIIA mutants and Tor-OE , in which case our ribosome profiling approach may have lacked sufficient resolution to detect these changes , as tagged ribosomes were purified from whole muscle lysates . Indeed , a recent report demonstrated synapse-specific PHP expression [58] . Future studies utilizing genetic , electrophysiological , biochemical , and imaging approaches will be necessary to identify the specific post-translational mechanisms that drive PHP signaling , and to what extent shared or distinct mechanisms are common between pharmacologic , genetic , and Tor-OE mediated PHP signaling . Cells possess a remarkable ability to homeostatically control protein expression and stability , a process called proteostasis [60] . This requires a robust and highly orchestrated balance between gene transcription , mRNA translation , and protein degradation [45 , 61] , while disruption of this process contributes to aging and disease [62 , 63] . Further , proteostatic mechanisms are not only customized to the unique demands of specific cells and tissues , but are adjusted throughout developmental stages and even tuned over hours according to diurnal metabolic and feeding cycles [64–66] . The homeostatic nature of proteostasis is highlighted by the adaptations triggered in response to perturbations that threaten stable cellular protein levels , such as starvation and inhibitions of protein degradation [67 , 68] . We have used transcriptional and ribosome profiling to reveal new homeostatic adaptations triggered by proteostatic mechanisms that stabilize the proteome following chronic elevations in protein synthesis . In particular , genes that promote protein stability ( chaperones ) , protein degradation , and ribosome biogenesis were transcriptionally and/or translationally upregulated following Tor overexpression in muscle ( Fig 7 ) , modulations in complementary pathways that synergistically prevent inappropriate protein interactions , promote protein removal , and increase the machinery necessary to maintain elevated protein synthesis [47 , 53 , 69] . Interestingly , many of these pathways are also targeted following other homeostatic perturbations to proteome stability , including heat shock , starvation , and inhibitions in protein degradation [67 , 70] . This may suggest that proteostatic signaling involves a core program orchestrating adaptive modulations to transcription and translation in response to a diverse set of challenges to protein stability . Thus , ribosome profiling enabled the definition of transcriptional and translational mechanisms that respond to chronic elevations of protein synthesis , revealing changes in translation that would not be apparent through profiling of total RNA expression alone . Recent developments in next-generation sequencing have greatly expanded our ability to investigate complex biological phenomena on genome-wide scales . The power and variety of sophisticated genetic approaches are well-known in Drosophila . These include tissue-specific expression with a broad array of Gal4 and LexA drivers , transposable element manipulations , CRISPR/Cas-9 gene editing , and extensive collections of genetic mutations and RNAi lines [71–74] . Although some approaches have emerged that permit the analysis of RNA from entire organs as well as ribosome-associated RNA from specific tissues [5–7 , 9 , 38 , 75 , 76] , the technology described here now adds ribosome profiling to join this powerful toolkit to enable the characterization of translational regulation in specific cells with unprecedented sensitivity .
Drosophila stocks were raised at 25°C on standard molasses food . The w1118 strain is used as the wild type control unless otherwise noted , as this is the genetic background of the transgenic lines and other genotypes used in this study . The following fly stocks were used: GluRIIASP16 [20] , UAS-Tor-myc [77] , RpL3G13893 ( Bloomington Drosophila Stock Center , BDSC , Bloomington , IN , USA ) , RpL3KG05440 ( BDSC ) . All other Drosophila stocks were obtained from the BDSC . To control for the effects of genetic background on next generation sequencing data , we generated an isogenic stock and bred the genetic elements used in this study , ( BG57-Gal4 , UAS-RpL3-Flag , GluRIIASP16 , and UAS-Tor-myc ) into this isogenic line by outcrossing for five generations to minimize differences in the genetic background . During initial testing phases to determine the optimal ribosome subunit and biochemical tag to use , we generated several constructs and systematically compared purification efficiency . In particular , we inserted 1xFlag-6XHis tags to the C-terminals of the ribosome subunits RpL3 , RpL36 , RpS12 , and RpS13 . We engineered expression with each subunit’s genomic promotor into the pattB vector [78] . Transgenic stocks were made and tested for affinity purification of intact ribosomes using cobalt ion-coupled beads ( Clontech , 635501 ) . These biochemical tags were found to be inferior when compared to a single 3xFlag tag , which was used for the design of all subsequent constructs . To generate the UAS-RpL3-3xFlag and UAS-RpS13-3xFlag transgenic lines , we obtained cDNA containing the entire coding sequences of RpL3 ( FBcl0179489 ) and RpS13 ( FBcl0171161 ) . RpL3 and RpS13 coding sequence were PCR amplified and sub-cloned into the pACU2 vector [31] with C-terminal 3xflag tag using a standard T4 DNA ligase based cloning strategy . To generate the genomic RpL3-3xflag construct , a 6 . 5kb sequence containing the entire RpL3 genomic locus was PCR amplified from a genomic DNA preparation of w1118 using the following primers 5’-ATCGGTACCACTTACTCCCTTGTTG-3’ and 5’-CAGCTGCAGGGTTTGTGACTGATCTAAAAG-3’ . The same linker-3xflag sequence used in UAS-RpL3-3xflag was inserted before the stop codon of RpL3 using extension PCR . This sequence was cloned into the pattB vector [78] . Constructs were sequence verified and sent to BestGene Inc . ( Chino Hills , CA ) for transgenic integration . All libraries were sequenced on the Illumina NextSeq platform ( single read , 75 cycles ) , and three replicates were performed for each genotype . All sequencing datasets are deposited in the NCBI GEO datasets , accession number: GSE99920 . Sequencing data analysis was performed using CLC genomics Workbench 8 . 0 software ( Qiagen ) . Raw reads were trimmed based on quality scores , and adaptor sequences were removed from reads . Trimmed high quality reads were then mapped to the Drosophila genome ( Drosophila melanogaster , NCBI genome release 5_48 ) . Only genes with more than 10 reads uniquely mapped to their exons were considered to be reliably detected and further analyzed , as the variability was sharply higher for genes with less than 10 mapped reads compared to genes with mapped reads above 10 ( S5 Fig ) . We excluded genes from further analysis that were only found to be transcriptionally expressed , which were likely to result from non-muscle RNA . Relative mRNA expression levels were quantified by calculating RPKM ( Reads Per Kilobase of exon per Million mapped reads ) using mapping results from transcriptional profiling . Relative translation levels were quantified by calculating RPKM using mapping results from ribosome profiling . Translation efficiency was calculated by dividing ribosome profiling ( or translational profiling TRAP ) RPKM by transcriptional profiling RPKM . To determine differentially transcribed or translated genes , a weighted t-type test [81] was performed based on three replicate expression values for each gene between GluRIIA mutants and wild type , and Tor-OE and wild type using the statistical analysis tool of CLC genomics workbench . The analysis was performed on expression values obtained by transcriptional profiling to determine differentially transcribed genes , and on expression values obtained by ribosome profiling to determine differentially translated genes . Genes with a p-value less than 0 . 05 and fold change higher than 3-fold were considered differentially transcribed or translated unless otherwise stated . We also determined differentially transcribed or translated genes using R package DESeq2 analysis [82] , considering genes with adjusted p-values less than 0 . 05 as differentially expressed . The Baggerly’s t test method and DESeq2 method produced highly similar lists of differentially expressed genes . To determine gene targets undergoing translational regulation in GluRIIA mutants and Tor-OE compared to wild type , two criteria were used . First , the gene must have at least a 2-fold significant increase ( p<0 . 05 , Student’s t test ) in translation efficiency compared to wild type . Second , a significant increase in ribosome profiling expression value ( p<0 . 05 , Baggerly’s t test ) must also exist for the same gene . These two criteria ensure identification of genes that have true translational up-regulation that are not due to transcriptional changes . Metagene analysis was performed using Plastid analysis software [83] using default settings . Third-instar larvae were dissected in ice cold 0 Ca2+ HL-3 and fixed in Bouin's fixative for 2 min and immunostained and imaged as described [84] . Quantitative PCR ( qPCR ) was performed using Luna® Universal One-Step RT-qPCR Kit ( NEB , E3005S ) according to manufacturer’s instructions . RNA was isolated and prepared from body wall tissue as described above . 5 ng of total RNA was used as template in each reaction . Three biological replicates were performed for each sample and the 2^-ΔΔCt method was used for qPCR data analysis . The primers used for assaying each target are as follows ( fwd/rev , 5’-3’ ) : All recordings were performed in modified HL-3 saline with 0 . 3 mM Ca2+ as described [85] . All data are presented as mean +/-SEM . A Student’s t test was used to compare two groups . A one-way ANOVA followed by a post-hoc Bonferroni’s test was used to compare three or more groups . All data was analyzed using Graphpad Prism or Microsoft Excel software , with varying levels of significance assessed as p<0 . 05 ( * ) , p<0 . 01 ( ** ) , p<0 . 001 ( *** ) , N . S . = not significant . Statistical analysis on next generation sequencing data was described in the High-throughput sequencing and data analysis section . | Recent advances in next-generation sequencing approaches have revolutionized our understanding of transcriptional expression in diverse systems . However , transcriptional expression alone does not necessarily report gene translation , the process of ultimate importance in understanding cellular function . Ribosome profiling is a powerful approach to quantify the number of ribosomes associated with each mRNA to determine rates of gene translation . However , ribosome profiling requires large quantities of starting material , limiting progress in developing tissue-specific approaches . Here , we have developed the first tissue-specific ribosome profiling system in Drosophila to reveal genome-wide changes in translation . We first demonstrate successful ribosome profiling from muscle cells that exhibit superior resolution compared to other translational profiling methods . We then use transcriptional and ribosome profiling to define whether transcriptional or translational mechanisms are necessary for synaptic signaling at the neuromuscular junction . Finally , we utilize ribosome profiling to reveal adaptive changes in cellular translation following cellular stress to muscle tissue . Together , this now enables the power of Drosophila genetics to be leveraged with ribosome profiling in specific tissues . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
]
| [
"invertebrates",
"gene",
"regulation",
"messenger",
"rna",
"animals",
"dna",
"transcription",
"animal",
"models",
"drosophila",
"melanogaster",
"model",
"organisms",
"protein",
"synthesis",
"experimental",
"organism",
"systems",
"cellular",
"structures",
"and",
"organelles",
"drosophila",
"chemical",
"synthesis",
"research",
"and",
"analysis",
"methods",
"proteins",
"gene",
"expression",
"insects",
"ribosomes",
"biosynthetic",
"techniques",
"arthropoda",
"biochemistry",
"rna",
"eukaryota",
"cell",
"biology",
"nucleic",
"acids",
"protein",
"translation",
"genetics",
"biology",
"and",
"life",
"sciences",
"organisms"
]
| 2017 | Development of a tissue-specific ribosome profiling approach in Drosophila enables genome-wide evaluation of translational adaptations |
Rickettsial infections and Q fever present similarly to other acute febrile illnesses , but are infrequently diagnosed because of limited diagnostic tools . Despite sporadic reports , rickettsial infections and Q fever have not been prospectively studied in Central America . We enrolled consecutive patients presenting with undifferentiated fever in western Nicaragua and collected epidemiologic and clinical data and acute and convalescent sera . We used ELISA for screening and paired sera to confirm acute ( ≥4-fold rise in titer ) spotted fever and typhus group rickettsial infections and Q fever as well as past ( stable titer ) infections . Characteristics associated with both acute and past infection were assessed . We enrolled 825 patients and identified acute rickettsial infections and acute Q fever in 0 . 9% and 1 . 3% , respectively . Clinical features were non-specific and neither rickettsial infections nor Q fever were considered or treated . Further study is warranted to define the burden of these infections in Central America .
Rickettsioses , including spotted fever group ( SFGR ) and typhus group ( TGR ) , and Q fever ( caused by Coxiella burnetii ) are increasingly recognized worldwide [1] . Both rickettsioses and Q fever often manifest as undifferentiated fever and cannot easily be distinguished clinically from other causes of acute febrile illness ( AFI ) . Furthermore , both are difficult to confirm in the laboratory , since convalescent sera , specific diagnostic reagents , and expertise are required . These infections are especially underappreciated in low resource settings where lack of laboratory capacity limits both individual diagnosis and validation of clinical acumen . Under-recognition may lead to unnecessary morbidity and even mortality , since empirical regimens for AFI usually do not treat rickettsioses or Q fever . In 1971 a large serosurvey documented the presence of rickettsiae and Q fever in humans in Central America [2] . However , these agents have not been prospectively studied in humans in Central America nor have there been cases reported of acute infection with these agents in Nicaragua . To identify , quantify , and characterize potentially treatable rickettsioses and Q fever among AFI in Nicaragua , we studied a cohort of children and adults presenting with fever at a large hospital .
Written informed consent was obtained from patients or their guardians ( for patients <18 years of age ) and written assent was obtained from patients aged 12–17 years . The institutional review boards of Johns Hopkins University and Duke University Medical Center ( USA ) as well as Universidad Nacional Autónoma de Nicaragua , León ( Nicaragua ) approved the study . We recruited patients in the emergency department and adult and pediatric wards of Hospital Escuela Oscar Danilo Rosales Arguello ( HEODRA ) , the 400-bed primary public teaching hospital of Universidad Nacional Autónoma de Nicaragua in León , Nicaragua , which serves rural areas around León as well as the city itself . Between August 2008 and May 2009 , we enrolled consecutive febrile ( ≥38°C , tympanic ) patients ≥1 month old without prior ( within 1 week ) trauma or hospitalization who presented during the day or early evening hours Monday through Saturday . Dedicated study doctors verified eligibility and willingness to return for follow-up and obtained written informed consent from patients ( ≥18 years ) or parents ( <18 years ) , and assent if 12–17 years . Study personnel recorded structured epidemiological and clinical data , including the duration of illness and clinical provider’s presumptive ( leading clinical ) diagnosis , on a standardized form at enrollment and then obtained peripheral blood specimens in EDTA and in a serum-separator tube for on-site clinician-requested testing and off-site research-related testing . Patients returned for clinical and serologic follow-up 2 to 4 weeks later , or were visited at home if they did not return and could be located . Blood in the serum separator tube was centrifuged and sera and EDTA blood frozen on site at -80°C . We used stringent criteria to define a confirmed case [6] . Confirmed acute rickettsial infection . ≥4 fold rise in IgG titer for SFGR and/or TGR; Confirmed past rickettsial infection . IgG titer by IFA of ≥160 for SFGR and/or TGR in the absence of a ≥4-fold rise in titer . Confirmed acute Q fever . ≥4 fold rise in C . burnetii Phase 2 IgG titer by IFA . Confirmed past Q fever . C . burnetii Phase 2 IgG titer of ≥160 by IFA in the absence of a ≥ 4-fold rise in titer . Probable acute rickettsial infection or Q fever . Seroconversion without a ≥4 fold rise in titer ( e . g . , acute-phase sera negative at 1:80 and convalescent sample positive at 1:80 . Probable past rickettsial infection or Q fever . IgG titer of 80 in the absence of seroconversion or 4-fold rise in titer . Those with equal SFGR and TGR titers were SFGR/TGR group-indeterminate . Possible acute rickettsial infection or Q fever . 2-fold rise in titer ( e . g . acute sample positive at 1:80 and convalescent sample positive at 1:160 ) SFGR vs . TGR infection . A ≥2-fold difference in titer defined SFGR vs . TGR; if titers were equal , the rickettsial infection was categorized as group indeterminate Acute phase EDTA-anticoagulated blood samples from patients with possible , probable , and confirmed acute rickettsial and/or Q fever infections ( IFA-confirmed seroconversion with convalescent titer ≥ 80 and/or a ≥2-fold rise in titer ) as well as past infections were used to prepare DNA for PCR analyses . For this , 1 mL of EDTA-anticoagulated blood was subjected to automated DNA preparation using a QIAsymphony SP instrument and the DSP DNA mini Kit; the final elution volume was 200 μL . DNA samples were initially tested using a multiplex 5’ nuclease assay targeting conserved regions in SFGR ompA , TGR 17 kDa genus-common antigen gene ( RURANT17KB ) , and human ACTB [7] . In addition , acute phase blood DNA samples from patients with a convalescent-phase SFGR and/or TGR IFA titer of ≥ 160 and a ≥ 2 fold increase in titer were tested separately for SFGR hypothetical protein gene RC0338 [8] , TGR hypothetical protein gene Rpr 274P [9] , and C . burnetii IS1111 spacer PCR ( courtesy D . Raoult , Unite des Rickettsies , Marseille , France ) [10] . We correlated epidemiologic and clinical findings with serologic results . Proportions were compared by the Chi-square test or Fisher’s exact test and continuous variables by Student’s t-test or analysis of variance if normally distributed and Wilcoxon-Mann-Whitney or Kruskal-Wallis test if not normally distributed . Analyses were done with Stata IC 11 . 0 ( StataCorp , College Station , TX ) .
Serologic testing for rickettsial infections and Q fever was completed for 800 ( 97 . 0% ) of 825 consecutively enrolled patients . Of these 800 , 748 ( 90 . 7% ) had paired sera available , since 52 patients did not return and could not be located for follow-up . The likelihood of a subject returning for convalescent serum sampling and clinical follow-up did not differ by age ( p = 0 . 90 ) , sex ( p = 0 . 93 ) , or self-reported urban vs . rural residence ( p = 0 . 53 ) . The reported median distance from residence to hospital was 2 km for those who followed up versus 3 km for those who did not ( p = 0 . 08 ) . Among the 748 patients with paired sera , the median age was 9 years ( IQR 3–29 ) . Slightly more were male ( 52 . 5% ) , and males were younger than females ( median age 9 vs . 11 , p = 0 . 007 ) . The median reported duration of fever and of illness at presentation was 2 days ( IQR 1–4 ) and 3 days ( IQR 1–5 ) , respectively . Many ( 30 . 0% ) reported taking an antibiotic before presentation . The median interval between acute and convalescent follow-up was 15 days ( IQR 14–28 ) . A total of 5 patients were treated with doxycycline , none of whom had acute rickettsial infection or acute Q fever . Overall , 14 ( 2% ) of 747 patients had evidence of Q fever infection . Patients with definitive serologic evidence of Q fever infection were older than those without ( median 36 vs . 9 years , p = 0 . 0003 ) ; 3 . 9% of adults age 18 years or older were seropositive for Q fever compared with 0 . 7% of children ( p = 0 . 002 ) . Acute Q fever occurred throughout the study period ( 1 each in October to March except for 3 in April and 2 in November ) without seasonality or specific association with rainfall or temperature . Otherwise , despite extensive investigation , no demographic characteristics or environmental exposures were associated with Q fever . However , patients with Q fever infection were more likely than others to have definitive serologic evidence of rickettsial infection ( 21% vs 4% , p = 0 . 002 ) .
Using stringent serodiagnostic criteria and achieving 90% follow-up , we document and describe Rickettsia spp . and C . burnetii as causes of acute febrile illness in Nicaragua . We required a 4-fold change in IgG titer to define a laboratory-confirmed case of acute rickettsial infection , which is in keeping with the latest ( 2010 ) case definition for acute spotted fever infection from the Centers for Disease and Prevention [6] . We found that SFGR and C . burnetii caused undifferentiated febrile illness predominantly in adults , especially male adults for SFGR . These infections mimicked other acute febrile illness and were both unsuspected and untreated , as we found in a similar cohort study in Sri Lanka [11] . Improved awareness and diagnostic tests may decrease morbidity and mortality by enhancing case detection and prompt provision of appropriate therapy . It is plausible that rickettsioses , both SFGR and TGR , would cause acute febrile illness in Nicaragua . R . rickettsii , the cause of Rocky Mountain spotted fever ( RMSF ) is distributed broadly throughout the Western Hemisphere , and confirmed cases have been documented elsewhere in Central America ( specifically Mexico , Costa Rica , and Panama ) [12–18] . Although not found in this study , fatal infections with R . rickettsii have been reported in Mexico , Costa Rica and Panama [15–17 , 19] , and fatal SFGR , presumably R . rickettsii , in Guatemala [14] . In Mexico , RMSF is directly linked to R . rickettsii-infected Rhipicephalus sanguineus ticks harbored by the large population of peridomestic stray dogs [20]; notably , in the U . S . , this same vector-host dynamic has been associated with 4-times higher RMSF case-fatality among American Indians compared with other ethnic groups [21–24] . Moreover , other important vectors of R . rickettsii in Central and South America ( Amblyomma mixtum and A . sculptum , respectively ) are present in Nicaragua , which heightens the likelihood that the serologic responses we observed are the result of R . rickettsii infection [25] . R . parkeri , another SFGR , is not yet implicated in human SFGR infection in Central America , but A . maculatum is present in the region and R . parkeri-like illness was reported in a traveler returning from Honduras [25 , 26] . In Brazil , Amblyomma ovale is the vector of Rickettsia sp . strain Atlantic rainforest , a R . parkeri-like agent [19] . Although serologic cross-reactions occur among species of SFG rickettsiae and serologic testing alone cannot directly identify a causative agent , higher titers in most cases to R . rickettsii than to R . parkeri suggest that R . rickettsii or a novel agent more closely related to R . rickettsii may have caused the confirmed SFG rickettsial infections . The relatively mild clinical illness we observed suggests the latter . Although C . burnetii is technically not a member of the Rickettsiales , the disease it causes , namely Q fever , is plausible as a cause of acute febrile illness in Nicaragua and is found worldwide when sought . Q fever was first identified in Central America ( Panama ) in the 1940s [27 , 28] and more recently a serosurvey of Q fever in livestock workers confirmed a seroprevalence as high as 10% [29] . A 1971 serosurvey of rickettsioses and Q fever in humans in Central America identified SFGR and Q fever antibodies by complement fixation and microagglutination and found an SFGR seroprevalence of 0 . 3% ( 1/312 ) and Q fever seroprevalence of 0 . 6%-1 . 0% [2] . We found that patients with acute rickettsial infections had illnesses and findings that closely resembled those of other patients with acute febrile illness . Patients with rickettsioses , however , were relatively older , reported a longer duration of fever , and more frequently had joint and muscle pain than did others; no patient had a rash . Although fever , headache , and rash constitute the classic clinical triad for rickettsioses , headache is frequent with other illnesses and rash is often absent when patients present early in illness , as in our cohort , or are individuals with dark skin [30] . Our group and others have also found that clinical characteristics are often not helpful in identifying rickettsial infections , with the exception possibly of older age [11 , 31–35] . In our study , acute rickettsial infections ( almost all SFGR ) were associated with self-reported rural residence , contact with livestock , and drinking well or river versus tap water . Tick-borne rickettsial infections are increasingly recognized as zoonotic infections that are more common in rural areas with complex tick-animal reservoir relationships that merit One Health approaches for control [11 , 36 , 37] . Patients with rickettsioses were also more likely to report exposure to ticks , fleas , and lice but not mosquitos . These findings are also plausible , since ticks and fleas harbor SFGR , and fleas and lice TGR . Patients with Q fever were also relatively older than other patients , but perhaps surprisingly did not report rural residence or exposure to livestock . We suspect this reflects small sample size , since individuals with confirmed acute Q fever did have acute Q fever-compatible illnesses , including hepatomegaly and prolonged illness associated with fever and cough . We found that both acute rickettsial infections and Q fever were unsuspected and untreated , despite the availability of doxycycline and its use in a few patients without rickettsioses . In addition to non-specific findings , rickettsioses and Q fever pose a diagnostic and therapeutic challenge because paired serology , the reference standard , is intrinsically retrospective and current PCR protocols are insensitive . Our ability to examine and identify distinguishing clinical features if present was greatly improved by the >90% follow-up albeit limited by sample size . Although many SFGR infections and Q fever cases are self-limited , unconfirmed and potentially fatal rickettsial infections could have occurred in the 10% lost to follow-up , the unknown number of individuals too sick or poor to reach the hospital , and among the 5 who reported doxycycline treatment ( since treatment can dampen the IgG immune response ) [38] . Severe disease and deaths due to confirmed and presumed RMSF have been reported in Central America [15–18] , [12–16] with case fatality up to 20% [39] . However , the diversity , relative frequency , and clinical spectrum of SFG rickettsioses is not known in Central America , nor are the determinants of disease severity among SFG rickettsioses Conservatively , 5% ( 34/748 ) of patients with AFI had definitive evidence of rickettsial infection and 2% ( 14/748 ) Q fever . Our estimate is conservative because our algorithm of confirming ELISA positives with reference standard IFA assured specificity but not sensitivity , the latter of which would have required performing IFA in all patients . We required an IFA titer of 160 to be positive because we preferred to underreport rather than falsely report Rickettsia spp . and C . burnetii as newly identified causes of acute febrile illness in Nicaragua . However , it is likely that these etiologic agents caused much more than 6% of acute febrile illnesses and that the seroprevalence of rickettsioses and Q fever together was at least 10% . Because IFA is inherently subjective and readings can vary by one dilution even among experts [30] , we conservatively required a convalescent titer of 160 to define confirmed infection as we have done previously [11] , including in the setting of seroconversion . Confirmed acute rickettsial infections required demonstration of a 4-fold rise in IgG by IFA on paired sera , which is in consistent with the US Centers for Disease Control and Prevention’s decision to no longer accept a single high IFA titer or a positive ELISA as confirmatory for R . rickettsii infection [34] . We ascribed infections to SFGR rather than TGR if the former had 2-fold higher titers as we have done previously [11] . Although cross-reactions occur between groups and especially species within groups , titers would be expected to be higher in the homologous group . In addition to our stringent case definitions , it is likely that cases were missed because of lack of access to care and possibly also due to early treatment or mortality . A major strength of our study is use of SFG and TG ELISA to identify probable rickettsial infections followed by SFG and TG IFA on paired specimens to confirm acute vs . past infections . In contrast with existing seroepidemiological studies of rickettsial infections that solely utilize ELISA performed on a single serum sample , we opted to use a two-tiered approach that leverages the strengths of ELISA methods ( high throughput and objective evaluation ) but retains IFA ( gold standard , but somewhat cumbersome and subjective ) to strengthen certainty of our data . Our approach is in keeping with the US Centers for Disease Control and Prevention’s latest case definition [6] that classifies ELISA-positive cases as probable , and requires a 4-fold rise in titer by IFA to confirm SFG infections . We additionally found that ELISA screening should include both SFG and TG antigen , since two of 6 patients diagnosed with acute spotted fever rickettsiosis would have been missed if simultaneous IFA screening for both typhus and spotted fever group rickettsiae was not conducted . It is possible that we missed additional acute rickettsial infections that would have been identified had we performed IFA on the full cohort; hence , because of ELISA’s imperfect sensitivity in addition to specificity , we would emphasize the importance of IFA and empiric treatment for management of patients with suspected rickettsial infections . Given the limitations of serology , we also sought to confirm cases by PCR; however , PCR on whole blood is intrinsically insensitive for rickettsioses , due to low levels of bacteremia and the intracellular location of these bacteria within endothelial cells [40 , 41] . PCR for Coxiella is also insensitive , especially among seropositive patients [42 , 43] . Therefore , it is not surprising that paired IFA-confirmed infections were not corroborated by PCR . In summary , we provide definitive evidence and a conservative estimate of unsuspected and untreated rickettsial infections and Q fever among patients with AFI in Nicaragua . A population-based longitudinal study with speciation of SFGR will be required to define the full clinical spectrum of R . rickettsii vs . other possible SFGR species and the case fatality rate of specific SFGR , TGR , and Q fever in Nicaragua . Better diagnostic tests , evaluated relative to gold standard paired serology such as we achieved here , and further epidemiologic study will be necessary to understand the biology of human rickettsioses and Q fever in Central America , to identify vector-host relationships , and to guide treatment and preventive measures . | Rickettsial infections and Q fever cause illness characterized by fever and non-specific symptoms and signs . Not only are these infections difficult to recognize , they are also difficult to diagnose because of limitations in existing tests for them . Despite sporadic reports , rickettsial infections and Q fever have not been prospectively studied in Central America . We enrolled consecutive patients presenting with undifferentiated fever in western Nicaragua and collected data regarding potential risk factors as well as symptoms and signs associated with the illnesses . Additionally , we collected blood samples at the initial visit and 2 to 4 weeks thereafter . We used serologic assays to differentiate new ( rising antibody titers ) vs . old ( stable antibody titers ) infections . Characteristics associated with both acute and past infection were assessed . We enrolled 825 patients and identified acute ( new ) rickettsial infections and acute Q fever in 0 . 9% and 1 . 3% , respectively . Clinical features were non-specific and neither rickettsial infections nor Q fever were considered nor treated . Further study is warranted to define the burden of these infections in Central America . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
]
| [
"medicine",
"and",
"health",
"sciences",
"enzyme-linked",
"immunoassays",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"geographical",
"locations",
"microbiology",
"rickettsia",
"north",
"america",
"bacterial",
"diseases",
"signs",
"and",
"symptoms",
"antibodies",
"immunologic",
"techniques",
"bacteria",
"bacterial",
"pathogens",
"research",
"and",
"analysis",
"methods",
"central",
"america",
"coxiella",
"burnetii",
"immune",
"system",
"proteins",
"infectious",
"diseases",
"rickettsia",
"rickettsii",
"serology",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"immunoassays",
"people",
"and",
"places",
"biochemistry",
"diagnostic",
"medicine",
"fevers",
"physiology",
"biology",
"and",
"life",
"sciences",
"q",
"fever",
"organisms"
]
| 2016 | First Identification and Description of Rickettsioses and Q Fever as Causes of Acute Febrile Illness in Nicaragua |
Most mammalian genes are able to express several splice variants in a phenomenon known as alternative splicing . Serious alterations of alternative splicing occur in cancer tissues , leading to expression of multiple aberrant splice forms . Most studies of alternative splicing defects have focused on the identification of cancer-specific splice variants as potential therapeutic targets . Here , we examine instead the bulk of non-specific transcript isoforms and analyze their level of disorder using a measure of uncertainty called Shannon's entropy . We compare isoform expression entropy in normal and cancer tissues from the same anatomical site for different classes of transcript variations: alternative splicing , polyadenylation , and transcription initiation . Whereas alternative initiation and polyadenylation show no significant gain or loss of entropy between normal and cancer tissues , alternative splicing shows highly significant entropy gains for 13 of the 27 cancers studied . This entropy gain is characterized by a flattening in the expression profile of normal isoforms and is correlated to the level of estimated cellular proliferation in the cancer tissue . Interestingly , the genes that present the highest entropy gain are enriched in splicing factors . We provide here the first quantitative estimate of splicing disruption in cancer . The expression of normal splice variants is widely and significantly disrupted in at least half of the cancers studied . We postulate that such splicing disorders may develop in part from splicing alteration in key splice factors , which in turn significantly impact multiple target genes .
The majority of mammalian genes produce alternative transcripts as part of their normal expression program [1]–[4] . Alternative transcripts include splicing , polyadenylation and transcription initiation variants which can be expressed differentially in different tissues [4]–[7] providing the fine tuning of gene expression required for cell differentiation and tissue-specific functions . Disruptions in the balance of alternative transcripts , especially at the splicing level , are known to affect angiogenesis [8] , cell differentiation [9] and invasion [10] . A large body of evidence has established connections between alternative splicing defects and cancer , so that the identification of transcript isoforms is now considered an important avenue in cancer diagnosis and therapy [11] , [12] . The disruption of splicing isoform expression in cancer may result from very different underlying genetic events . On one hand , mutations in cis-regulatory sequences lead to the abnormal expression of specific isoforms , as observed for example in the BRCA1 gene in breast and ovarian cancer [13] . Another class of event includes alterations of the mRNA processing machinery or its signalling pathway . These may affect the splicing of specific genes such as CD44 [14]–[16] , but may also cause wider perturbations of isoform expression as the processing of multiple genes can be simultaneously affected [17]–[20] . Evidence for wider changes in alternative transcription linked with cancer are present for instance in EST databases , where a large fraction of splice variant are actually tumor-specific [21] . However , while most studies of splicing and cancer attempt to isolate “signature” splice variants with significant over-expression in disease cells , no published work to date has focused on the bulk of splicing disruption that potentially arises when the splicing machinery is impaired . The aim of the present study is to evaluate the extent and modalities of non-specific alternative transcript disruptions in cancer . Instead of seeking “interesting” signature isoforms , we analyzed the distribution of all isoforms from a single gene in a given tissue . We postulated that , in a tissue where the splicing machinery is impaired , the distribution of isoforms may be more disordered than in a control tissue . To measure the level of disorder in cDNA and cDNA tag libraries , we borrowed the notion of entropy from information theory . We applied this measure to all three types of alternative transcription , comparing isoform distributions in pairs of disease and normal tissues . Our results show that neither alternative polyadenylation nor alternative transcription initiation are associated with a disordered isoform expression . However , in half of the cancers studied , alternative splicing showed a highly significant entropy gain relative to the corresponding normal tissues . We analyze this entropy gain and discuss its possible causes .
Given a random variable X with probabilities P ( xi ) for discrete set of events x1 , … . , k , Shannon's entropy , also known as Information Entropy , is defined by:The entropy , and thus the disorder , is maximal when the probability of all the events P ( xi ) are equal and thus the outcome is most uncertain . Here , Shannon's entropy is applied to the expression profiles of different transcript isoforms for a given context . In the Figure 1 example , Gene1 has 4 alternative splice forms ( SP1…SP4 ) and we are interested in their expression in normal cerebellum and cerebellum tumor tissues . For each splice form , we count the number of transcripts observed in different tissue types ( for instance ESTs/cDNAs matching splice form SP1 are observed 4 times in cerebellum tumor libraries and once in normal tissue libraries ) . For this gene , isoform entropy across the four splice forms is higher in tumor than in normal cerebellum tissues , reflecting a more uniform tissue distribution of isoforms in the tumor libraries . We hypothesised that impairment of the transcriptional or post-transcriptional control machinery in cancer or other diseases should result in the loss of a tissue-specific expression pattern of certain transcript isoforms . This loss can be measured by a gain of entropy in the expression pattern of isoforms of a given gene . By averaging entropy gains or losses on a sufficient number of genes expressed in a disease/normal tissue pair , we should observe a significant entropy bias if isoform expression is altered in this disease . We obtained transcript isoform collections from the FANTOM3 database [1] for initiation variants and the ATD database [22] for polyadenylation and splicing variants . We then related isoforms to cDNA or cDNA tag counts and mapped each cDNA or tag to its tissue/disease information using the EvoC ontology [23] for ESTs/cDNAs or direct parsing of CAGE/SAGE databases as explained in Materials and Methods . A gene was considered in the entropy calculation only if it had at least two alternative isoforms supported by at least 10 different transcripts from three separate libraries , thus a total of at least 20 transcripts mapped to each gene considered . In order to measure isoform entropy changes in a disease/normal tissue pair , we required that at least 50 genes and 100 isoforms were found expressed in both the normal and disease tissues . By considering only isoforms that were observed in both states , we excluded from our analysis spurious isoforms that are prevalent in many cancer EST libraries [24] . We define the entropy ratio of a gene as the ratio of the entropy of this gene in the disease to the entropy of the same gene in the normal tissue . The entropy ratio of a disease/normal tissue pair is the average of the entropy ratios of all genes available in this tissue pair . Figure 2 presents entropy ratios for different diseases with respect to alternative initiation ( A ) , polyadenylation ( B ) and splicing ( C ) . An entropy ratio of one means that isoform entropy does not vary between disease and normal tissue ( thick line in Figure 2 ) . To estimate significance boundaries , random assays were performed by dividing the average entropy of 1000 randomly picked genes from any disease/tissue state by that of another randomly picked set of 1000 genes from any other disease/tissue state and repeating this process 10 , 000 times . This process was performed independently on the three isoform datasets . Values for the highest and lowest percentile are represented by red and green vertical lines , respectively . Entropy ratios for alternative initiation and polyadenylation did not ever exceed the significance boundaries ( Figure 2A and 2B ) in the 6+8 cancer/normal tissue pair studied . This suggests that expression of alternative polyadenylation and initiation isoforms does not present large scale alterations in cancer . Alternative splicing however was quite different with 24 of the 27 cancer tissues studied showing a higher level of entropy than their normal counterpart ( Figure 2C and Table S1 ) . This entropy gain was highly significant in 13 cases , suggesting that the expression of splicing isoforms is strongly disrupted in certain cancers . In none of the 27 cases studied did the normal tissues show significantly higher entropy than disease tissues , and none of the three non-cancer diseases ( arthritis , ascites and schizophrenia ) presented a significant entropy change between normal and disease tissues . The observed entropy bias is not imputable to sampling differences in normal and cancer libraries . The number of ESTs/cDNAs used to calculate entropy did not differ significantly between normal or disease tissues ( Table S1 ) , mainly due to the fact that we considered only isoforms that are expressed both in disease and normal tissues . Furthermore , Pearson's correlation tests ( Table S1 ) showed no relationship between the entropy ratio and differences in the numbers of ESTs/cDNAs between normal and disease tissues ( P = 0 . 28 ) or between the entropy ratio and the total size of libraries ( P = 0 . 12 ) . The observed gain in entropy can therefore not be attributed to a size effect of cancer EST libraries . In the ten most disrupted cancer tissues , splicing entropy gains were caused by 16 to 258 significantly disrupted genes , or 30%–68% of the gene set available for entropy calculation in these tissues . This suggests that splicing perturbation is caused by factors that regulate multiple genes at the same time . Sets of splice-disrupted genes from different tissues show little overlap therefore we cannot isolate a list of genes displaying a generally higher rate of splicing disruption . However , a clear functional trend appears when high entropy gain tissues are pooled together . In the ten cancer tissues that displayed the highest gain in splicing entropy ( from stomach/carcinoma to brain/astrocytoma , Figure 2 ) , we analyzed all genes showing a splicing entropy gain ( 414 genes ) for functional enrichment . Interestingly , the most over-represented terms among splice-disrupted genes either contain “RNA splicing” or are higher level terms that incorporate RNA splicing ( Table 1 ) . The “RNA splicing” class mostly comprises splice factors . This suggests that splicing alterations in a few key splice factors could be involved in the more extensive splicing disruption observed in the high entropy-gain tissues . This enrichment is observable only after cancer tissues are pooled , which means the number of disrupted splice factors in a single disease is low . A total of 13 splice factors show a significant increase in splicing entropy in the cancer tissues studied ( Table S2 ) . Most are constitutive splice factors , only three ( TRA2B , U2AF1 , SF3A2 ) being involved in alternative splicing regulation . Splice factors are subject to alternative splicing at higher rates than average genes: 72% of the 58 annotated splice factors in Gene Ontology [22] have at least one alternative splice form in the ATD database [25] , with an average of 5 . 4 isoform per gene , compared to 62% alternative splicing and 3 . 4 isoform per gene in the total ATD gene set . To test whether this bias could explain the over-representation of splice factors among disrupted genes in the high entropy gain cancers , we performed the same GO-term analysis among splice-disrupted genes in the ten disease categories displaying the lowest entropy gain . We could not observe any functional bias in this gene set ( not shown ) . Therefore , splicing deregulation of splice factors is a hallmark of tissues where overall splicing is deregulated . This again designates misplicing of splice factors as a possible cause of wider splicing disruption in these tissues . Although tumors are diverse and heterogeneous , they all share the key ability to proliferate at a higher level than normal tissue and this despite the very tight control that the organism usually exerts on cell proliferation . To test potential links between disordered isoform expression and higher levels of proliferation , we classified the cancer types that deregulate the splicing mechanism ( Figure 2C ) in function of their proliferative potential . To evaluate proliferation , we extracted the 188 genes from the “cell cycle” module of Stuart et al . [26] , a cluster of coexpressed genes shown to be enriched in elements that are overexpressed in highly proliferative cells and whose high expression is a marker of entry into the cell cycle [27] . We manually verified each of these 188 genes ( Table S3 ) and confirmed that 92 were shown to be specifically over-expressed during one of the replicative phases of the cell cycle and another 17 bore significant proof of being over-expressed in proliferating cells . We thus used a high expression of these markers as a surrogate for a high level of proliferation . In order to obtain a “proliferation index” of cancer samples , we computed the median expression level of the 188 markers in each of 3787 published Affymetrix microarray experiments performed on cancer samples [28] . Samples were then binned into five categories from low to high proliferation , as shown in Figure 3 . To relate proliferation levels to splicing entropy results , we considered only microarray samples that contained the exact same keywords as disease tissues in Figure 2C . Results are shown in Figure 4 . Cell proliferation , as measured from the expression of cell cycle genes , is significantly correlated to splicing entropy gains . This observation led us to question the possible correlation between splicing entropy and cellular proliferation in a non-pathological context . We compared the splice isoform entropy of foetal and adult tissues in the same manner we compared disease and normal tissues ( Figure 5 ) . While foetal tissues are expected to present higher levels of proliferation than their adult counterparts , we could not observe any significant entropy gain in foetal tissues . This suggests the higher isoform entropy observed in highly proliferating cancers is only indirectly related to proliferation ( proliferation indices of foetal tissues could not be obtained due to insufficient foetal microarray data ) .
While previous studies of cancer-related splicing alterations have focused mainly on the discovery of “aberrant” splice variants , we looked instead at changes in the balance of variants expressed in both healthy and cancer tissues . This new perspective enabled us to characterize another kind of splicing disorder in which splice variant expression profiles are significantly flattened in tumors . While isoforms from the same gene are usually differentially expressed in a given tissue , with clear minor and major forms , these expression differences are reduced in cancer and this leads to a raise of isoform entropy . Although controlled over/under-expression events may in principle produce a flattened profile , we find unlikely that the generalized entropy gain observed in cancer could result from a combination of multiple controlled changes in isoform expression . The entropy gain is more likely a sign of a general loss of regulation involving widespread , non-specific perturbations of alternative splicing . We did not observe such cancer-related disorders in alternative transcription initiation and alternative polyadenylation , the two other processes associated with expression of disease-specific isoforms . Previous efforts to identify cancer-specific splice forms , either through EST analysis or experimental means , have mostly ignored non-specific , large-scale disruptions . An exception is the study by Xu and Lee [29] which sought splice forms with statistically significant expression changes between normal and tumor EST libraries . In that sense , these authors were looking for events that would cause an entropy reduction , not an entropy gain . However , they also discussed the impact of unspecific disruptions and analyzed expression patterns that may lead to cancer-specific isoforms ( Figure 6 ) . The most frequent patterns leading to cancer-specific events were the loss of a normal isoform S , and the switch in expression between normal ( S ) and cancer-specific ( S' ) isoforms . A general entropy gain would go against the occurrence of such events , which makes these patterns even more interesting on a background of entropy gain . Contrarily , the “gain of S'” category is directly correlated to a rise of entropy ( i . e . the “tumor” situation has higher entropy ) . Therefore , in a context of general entropy gain , events of the “gain of S'” category , even when statistically significant , could merely reflect the wider splicing disruption and should be considered with caution . Xu and Lee rightly noted that this category , which produces only a small fraction of cancer-specific splice forms , may be related to a loss of splicing specificity in tumors . There is now ample evidence that changes in splice factor expression , due for instance to kinase activation [14] , disrupt splicing patterns in tumors [16] , [18]–[20] , [30] , [31] . Figure 7 , box A presents the most common of these effects , where an up-regulated splice factor causes expression of a rare or aberrant splice form . Splice factors previously analyzed for such dysfunctions include SF2/ASF , U2AF-65 , SFRS2 , SFRS3 , SRm160 , hnRNP A1/A2 , and TRA2-β , all acting both in alternative and constitutive splicing . Although these factors may potentially target many genes , studies have focused on specific targets such as CD44 and have not examined more widespread splice defects . The splicing disruptions that we observed apparently affect a larger number of transcripts and are characterized by a loss of splice form regulation . Although this phenomenon might occur as a byproduct of the above mechanism , its association with the mis-splicing of splice factors , prevalently of the constitutive type , leads us to postulate a second process ( Figure 7 , box B ) in which mis-splicing of general splice factors would cascade into a wider splicing disruption and entropy gains . Among the 13 splice factors that displayed splicing disruptions in our study , two were already known to regulate their own splicing: SFRS3 and TRA2-β [15] , [28] . In each case , overexpression of the splice factor activated the inclusion of stop codon-containing exons [15] , [28] producing transcripts subject to nonsense-mediated decay [32] , [33] . Both genes have additional isoforms that are not NMD-prone ( Figure S1 ) and may contribute to the mis-splicing of other genes . A possible link between the two pathways in Figure 7 naturally comes to mind when considering that a change in splice factor expression in pathway “A” could alter the splice variant balance of other splice factors in pathway “B” . This transition may occur preferentially in highly proliferating tumors , where we observed the strongest splicing disruption . Splicing perturbation is knowingly correlated to proliferation [31] however no causal relationship between these events has been identified yet . Perhaps the splicing mechanism has trouble in trying to keep up with the accelerated pace of cell proliferation or a general disorder in splicing is causing failure in the regulation of cell cycle . Independently of any mechanistic hypothesis , splicing entropy measures show that widespread splicing disruption may be prevalent in most cancer tissues . In such a context of high splicing entropy , therapeutic avenues involving the reprogrammation of mis-spliced isoforms [34] would have a limited interest . As already recognized in different studies [35] , [36] splice factors or their regulatory machinery may turn out as better therapeutic targets .
Transcripts and expression data for each type of transcriptional variation ( initiation , splicing , polyadenylation ) were obtained from the following sources . Alternative initiation isoforms were obtained from the CAGE Basic/Analysis databases at http://fantom31p . gsc . riken . jp/cage_analysis/hg17/ . This database classifies 3 , 106 , 472 CAGE tags into 450 , 228 transcription clusters ( TC ) further grouped into 32 , 351 transcription units ( TU ) . TCs and TUs are two operationally defined units proposed in FANTOM3 [1] used to characterize promoters and genes respectively . We considered only those TCs that bore proof from at least 3 different CAGE libraries and 10 transcripts . These TCs were downloaded from the RIKEN website as well as the mappings of CAGE transcripts to these TCs in a given tissue type . This allowed us to create a relational database in which each TC could be queried to display its mapped CAGEs in each tissue type and the TU to which it belongs . For each normal/disease tissue pair we could therefore query a list of TCs common to both tissue types , link these TCs to their specific TUs and obtain the number of CAGEs mapped to a each of these TCs from the normal tissue library and from the disease tissue library . Alternative polyadenylation isoforms were downloaded from the EBI ATD database , Human Release 1 ( 31 May 2005 ) [25] at http://www . ebi . ac . uk/atd/humrel1 . html . Here , we only considered poly ( A ) sites located in the 3′-most exon of the gene because poly ( A ) sites located in upstream exons can belong to different splice forms . Since alternative splicing and polyadenylation can interfere [37] , such events cannot be safely attributed to either phenomena . Again , each alternative polyadenylation event had to be supported by three different cDNA libraries and 10 transcripts , giving a total of 206 , 138 transcripts mapped to 13 , 367 poly ( A ) sites for 4400 genes . These 13 , 367 poly ( A ) sites were downloaded from the ATD website as well as the mapping of ESTs , cDNAs and SAGES to these isoforms . cDNA and EST transcripts were then linked to the eVOC 2 . 6 ontology through their Genbank accession identifiers and SAGE transcripts were manually parsed for simple tissue descriptors that were identical to eVOC 2 . 6 ontology terms ( 39 descriptors from the Gene Expression Omnibus [27] ) . This allowed us to create a relational database in which each poly ( A ) isoform could be queried to display its mapped transcripts in each tissue type and the Ensembl gene ID to which it belonged . For each normal/disease tissue pair we could therefore query a list of poly ( A ) isoforms common to both tissue types , link these isoforms to their specific Ensembl gene identifier and obtain the number of transcripts mapped to a each of these isoforms from the normal tissue library and from the disease tissue library . Alternate splice isoforms were also downloaded from the EBI ATD database , Human Release 1 . Again , 3 separate libraries and 10 transcripts were required to establish a splice form . Transcripts that mapped to multiple isoforms were excluded from the study bringing the total number of transcripts/isoforms/genes in the database from 808845 / 52742 / 14791 to 444799 / 47308 / 12281 . These 47 , 308 alternative splice sites were downloaded from the ATD website as well as the mapping of ESTs and cDNAs to these isoforms . cDNA and EST transcripts were then linked to the eVOC 2 . 6 ontology through their Genbank accession identifiers . This allowed us to create a relational database in which each alternative splicing isoform could be queried to display its mapped transcripts in each tissue type and the Ensembl gene ID to which it belonged . For each normal/disease tissue pair we could therefore query a list of splicing isoforms common to both tissue types , link these isoforms to their specific Ensembl gene identifier and obtain the number of transcripts mapped to a each of these isoforms from the normal tissue library and from the disease tissue library . Cell-cycle specific genes were extracted from the conserved co-expression network defined by Stuart et al . [26] and available for download at http://cmgm . stanford . edu/kimlab/multispecies . A matrix of gene-gene Euclidean distances was computed and used for hierarchical clustering using R software . The tree obtained was then split into several groups by specifying a cutoff height of 10 . All genes in the “cell cycle” cluster were extracted and their respective Locuslink ID used for annotation . Microarray expression data was obtained from the Gene Expression Omnibus [28] selecting Affymetrix GPL96 platform ( 8340 different samples ) . We parsed microarray sample descriptions for the presence of any EvoC ontology keyword inherited from the top level term ≪neoplasia≫ and then manually checked to see if the description genuinely corresponded to a cancer-related experiment . From a set of 8340 microarray samples studied , 3787 samples corresponded to cancer-related microarray experiments . Proliferation categories were then attributed to each sample based on the median ranking ( MR ) of the expression level of the 188 genes from the cell cycle node , as follows: High proliferation : MR in the top 20% of the genes on array . ; Medium-high proliferation : MR between top 20% and top 40% of genes on array; Medium proliferation : MR between the top 40% and top 60% of the genes on array; Medium-low proliferation: MR between bottom 20% and bottom 40% of genes on array; Low proliferation: MR in the bottom 20% of genes on array . | RNA splicing is the process by which gene products are pieced together to form a mature messenger RNA ( mRNA ) . In normal cells , RNA splicing is a tightly controlled process that leads to production of a well-defined set of mRNAs . Cancer cells , however , often produce aberrant , mis-spliced mRNAs . Such disorders have not been quantified to date . To this end , we use a well-known measure of disorder called Shannon's entropy . We show that overall splicing disorders are highly significant in many cancers , and that the extent of disorder may be correlated to the level of cell proliferation in each tumor . Surprisingly , genes that control the splicing mechanism are unusually frequent among genes affected by splicing disorders . This suggests that cancer cells may withstand harmful chain reactions in which splicing defects in key regulatory genes would in turn cause extensive splicing damage . As mis-spliced mRNAs are widely studied for cancer diagnosis , awareness of these global disorders is important to distinguish reliable cancer markers from background noise . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"computational",
"biology/alternative",
"splicing",
"genetics",
"and",
"genomics/gene",
"expression"
]
| 2008 | Entropy Measures Quantify Global Splicing Disorders in Cancer |
Base J , β-D-glucosyl-hydroxymethyluracil , is a chromatin modification of thymine in the nuclear DNA of flagellated protozoa of the order Kinetoplastida . In Trypanosoma brucei , J is enriched , along with histone H3 variant ( H3 . V ) , at sites involved in RNA Polymerase ( RNAP ) II termination and telomeric sites involved in regulating variant surface glycoprotein gene ( VSG ) transcription by RNAP I . Reduction of J in T . brucei indicated a role of J in the regulation of RNAP II termination , where the loss of J at specific sites within polycistronic gene clusters led to read-through transcription and increased expression of downstream genes . We now demonstrate that the loss of H3 . V leads to similar defects in RNAP II termination within gene clusters and increased expression of downstream genes . Gene derepression is intensified upon the subsequent loss of J in the H3 . V knockout . mRNA-seq indicates gene derepression includes VSG genes within the silent RNAP I transcribed telomeric gene clusters , suggesting an important role for H3 . V in telomeric gene repression and antigenic variation . Furthermore , the loss of H3 . V at regions of overlapping transcription at the end of convergent gene clusters leads to increased nascent RNA and siRNA production . Our results suggest base J and H3 . V can act independently as well as synergistically to regulate transcription termination and expression of coding and non-coding RNAs in T . brucei , depending on chromatin context ( and transcribing polymerase ) . As such these studies provide the first direct evidence for histone H3 . V negatively influencing transcription elongation to promote termination .
Kinetoplastids are early-diverged protozoa that include the human parasites Trypanosoma brucei , Trypanosoma cruzi , and Leishmania major , which cause African sleeping sickness , Chagas disease , and leishmaniasis , respectively . The genomes of kinetoplastids are arranged into long gene clusters , or polycistronic transcription units ( PTUs ) , which are transcribed by RNA polymerase ( RNAP ) II [1–3] . RNAP II transcription initiation and termination occurs at regions flanking PTUs called divergent strand switch regions ( dSSRs ) and convergent strand switch regions ( cSSRs ) , respectively [4] . Pre-messenger RNAs ( mRNA ) are processed to mature mRNA with the addition of a 5’ spliced leader sequence through trans-splicing , followed by 3’ polyadenylation [5–10] . The arrangement of genes into PTUs has led to the assumption that transcription is an unregulated process in these eukaryotes and a model in which gene regulation occurs strictly post-transcriptionally [11 , 12] . However , specific chromatin marks have been characterized at sites of transcription initiation and termination , including histone variants and modified DNA base J , which could function to regulate polycistronic transcription and gene expression [13–17] . Base J , β-D-glucosyl-hydroxymethyluracil , is a modified DNA base consisting of O-linked glycosylation of thymine in the genome of kinetoplastids and closely related unicellular flagellates [18 , 19] . Whilst J is largely a telomeric modification , it is also found internally within chromosomes at RNAP II transcription initiation and termination sites [13 , 20–25] . As reviewed in Borst and Sabatini ( 2008 ) , analysis of RNAP I transcribed telomeric polycistronic units in T . brucei led to the discovery of base J [20 , 26] . Regulation of the ~15 telomeric variant surface glycoprotein expression sites ( VSG ESs ) allows the parasite to evade the host immune system in a process called antigenic variation [27 , 28] . Although the genome of T . brucei has over 1 , 000 VSG genes , only one VSG is expressed at a given time . This is achieved through regulated transcription of the telomeric ESs , only one of which is productively transcribed at any time . The association of the modified base with silent ESs in the bloodstream life-cycle stage of the parasite has led to the hypothesis that J plays a role in the regulation of antigenic variation . However , no direct evidence has been provided . Base J is synthesized in a two-step pathway in which a thymidine hydroxylase , JBP1 or JBP2 , hydroxylates thymidine residues at specific positions in DNA to form hydroxymethyluracil , followed by the transfer of glucose to hydroxymethyluracil by the glucosyltransferase , JGT [26 , 29 , 30] . JBP1 and JBP2 belong to the TET/JBP subfamily of dioxygenases , which require Fe2+ and 2-oxoglutarate for activity [31–34] . The synthesis of base J can be inhibited by competitive inhibition of the thymidine hydroxylase domain of JBP1 and JBP2 by dimethyloxalylglycine ( DMOG ) , a structural analog of 2-oxoglutarate [31 , 35] . Removal of both JBP1 and JBP2 or the JGT also results in T . brucei cells devoid of base J [29–31 , 36] . The co-localization of base J with modified and variant histones at dSSRs and cSSRs suggested a functional role of modified DNA in the regulation of RNAP II transcription [13] . Our work in T . cruzi described a unique role of J in regulating RNAP II transcription initiation , where the loss of base J resulted in the formation of more active chromatin , increased RNAP II recruitment and increased PTU transcription rate [24 , 37] . Recent studies have described a role for base J regulating RNAP II termination in T . brucei and Leishmania . van Luenen et al . ( 2012 ) found that reduction of base J in L . tarentolae is associated with the generation of RNAs downstream of the cSSR that are antisense to the genes on the opposing gene cluster [25] . Reduction of base J in L . major resulted in similar defects [35] . Strand-specific RT-PCR detection of the nascent transcript confirmed that the J-dependent generation of RNAs downstream of the cSSR is due to read-through transcription at cSSR termination sites . In contrast , loss of J in T . brucei failed to indicate any defect in termination at cSSRs [35] . However , we localized base J at sites within PTUs where the loss of J led to read-through transcription and upregulated expression of downstream genes . Therefore , base J is required for RNAP II termination in both Leishmania and T . brucei , but to different degrees and at different locations . In L . major , J regulates termination at the end of each PTU to prevent read-through transcription and the generation of RNAs antisense to the genes on the opposing PTU . In contrast , although termination occurs at the end of each PTU in T . brucei in a J-independent manner , J-dependent termination within a PTU allows developmentally regulated expression of downstream genes . The core histones H2A , H2B , H3 and H4 , package DNA into nucleosomes and represent a critical component of higher order chromatin . All core histones have variant counterparts . Although histone post-translational modifications ( PTMs ) and their impact on transcription have been well documented , less is known about the role of histone variants in the regulation of transcription [38] . The most understood are variants of H2A and H3 . Several variants of H2A exist , including H2A . Z , H2A . B , and macroH2A . Both H2A . Z and H2A . B are associated with transcriptional activation [39–41] . Knockdown of H2A . Z inhibits transcriptional activation [42–44] . Consistent with this , and perhaps the most direct evidence of a transcriptional role of a histone variant , H2A . Z positively correlates with rates of RNAP II elongation , such that the reduction of H2A . Z increases RNAP II stalling [40] . Presumably , the nucleosome destabilizing effect of H2A . Z [45] leads to more accessible DNA at promoter regions for transcription factor binding , as well as promoting RNAP II elongation through gene bodies . Like H2A . Z , H2A . B is enriched at promoter regions and its reduction largely results in the downregulation of gene expression [39 , 46] . In contrast , macroH2A is enriched at transcriptionally repressed regions [47] and its reduction results in increased gene expression in an unknown mechanism [48] . Several H3 variants have also been characterized , including H3 . 3 and CENP-A , both of which are found in most eukaryotes including plants , mammals , and yeast . H3 . 3 differs from canonical H3 by only 4–5 amino acids and is found predominately at actively transcribed genes , forming more accessible nucleosomes [49–51] . H3 . 3 also provides a genome stabilization function at repetitive regions such as telomeres and centromeres [49 , 52–55] . Recent studies have implicated H3 . 3 in the maintenance of a repressed chromatin structure [56–58] . Evidence in mouse embryonic stem cells indicates H3 . 3 is enriched at lowly transcribed developmentally regulated genes where it promotes polycomb repressive complex 2 activity , which catalyzes the formation of the repressive modification H3K27me3 [57 , 58] . These findings suggest H3 . 3 maintains the promoters of developmentally regulated genes in a repressed , but transcriptionally “poised” state important for proper differentiation . H3 . 3 has also been implicated in the maintenance of H3K9me3 at endogenous retroviral elements in mouse embryonic stem cells [56] . Presence of H3 . 3 ( and H3K9me3 ) at endogenous retroviral elements repressed retrotransposition and expression of adjacent genes [56] . The centromeric specific histone variant has a well-characterized role in kinetochore formation , but its role in the regulation of transcription , if any , remains unknown [59] . Overall , although much progress has been achieved in the characterization of histone variants , few studies have revealed a direct link between histone variant function and transcriptional regulation . The J-independent nature of termination at cSSRs in T . brucei led us to characterize the role of H3 . V in regulating RNAP II termination . H3 . V and base J co-localize at RNAP II termination sites in T . brucei , including cSSRs and PTU internal termination sites [13 , 14] . H3 . V and base J also co-localize at telomeric repeats involved in regulating RNAP I transcription of the VSG expression sites [14 , 60] . T . brucei H3 . V shares 45% sequence identity with canonical H3 , much of the sequence divergence lying within the N-terminus , outside of the histone fold domain . H3 . V appears to be unique to kinetoplastids [14 , 60] and aside from its localization to termination sites and telomeres , very little is known about H3 . V and its potential role in the regulation of transcription termination . We demonstrate here that , similar to phenotypes associated with the loss of J , loss of H3 . V leads to defects in RNAP II termination within gene clusters and increased expression of downstream genes . Interestingly , many of the gene expression changes in the H3 . V knockout ( KO ) are further increased upon the subsequent loss of base J , suggesting that J and H3 . V have independent but overlapping roles in regulating transcription termination in T . brucei . Although the loss of H3 . V from cSSRs did not indicate any termination defects leading to transcription of the opposing strand of the adjacent convergent gene cluster , it does lead to increased generation of small interfering RNAs ( siRNAs ) that map to regions of overlapping transcription . Analysis of nascent RNA suggests this is due to increased transcription of the dual strand siRNA loci at cSSRs . We also detect increased expression of VSGs from silent VSG ESs in the H3 . V KO , indicating H3 . V can act independently in regulating telomeric repression and antigenic variation . Overall these findings provide the first known example of a histone H3 variant that functions as a repressive chromatin mark to promote transcription termination , in this case repressing both mRNAs and non-coding RNAs .
The co-localization of base J and H3 . V at RNAP II termination sites in T . brucei prompted us to examine the role of H3 . V in transcription termination . High-throughput sequencing of small RNAs has been shown previously to reveal transcription termination sites in trypanosomatids , as reflected in RNA degradation products [25 , 35] . The reduction of base J in L . major by treatment with DMOG resulted in the production of antisense small RNAs corresponding to genes in the opposing PTU due to read-through transcription at cSSRs [35] . In contrast , we found no evidence of termination defects in T . brucei at cSSRs following DMOG treatment and the complete loss of J . Antisense small RNAs , indicative of read-through transcription at cSSRs into the downstream PTU , were not increased following the loss of J [35] . We now show that the loss of H3 . V also does not result in read-through transcription at cSSRs . No significant changes in antisense small RNAs corresponding to read-through transcription at cSSRs into the downstream PTU were detected by small RNA-seq in the H3 . V KO compared to wild type ( WT ) T . brucei ( Fig 1A ) ( small RNA sequencing data discussed in this publication have been deposited in NCBI’s Gene Expression Omnibus and are accessible through GEO Series accession number GSE70229 ) . Subsequent loss of base J in the H3 . V KO parasites , via DMOG treatment , also failed to uncover any defect . These results suggest H3 . V is not required to prevent RNAP II read-through transcription at the end of convergent gene arrays in T . brucei . Unique to T . brucei , large peaks of sense and antisense small RNAs map to regions of overlapping transcription at the ends of convergent PTUs at cSSRs ( [35]; Fig 1A ) . These presumably represent the previously characterized Dicer 2-dependent Argonaute-associated siRNAs derived from cSSRs [61] . Consistent with this , we show that the RNAs that map to these regions correspond to the previously characterized siRNA size range in bloodstream form T . brucei ( 21-27nt ) [61–63] and exhibit characteristic phasing of siRNAs biogenesis , as indicated by the mapping of siRNA sequences in phased intervals at the target loci [64–66] ( Fig 1B and S1 Fig ) . This pattern suggests that these siRNAs were enzymatically processed , most likely by the DCR complex [67] , as opposed to random RNA degradation . siRNAs that map to cSSRs are 21-27nt compared to the 18-30nt range of small RNAs genome-wide ( Fig 1B and 1C ) . Although the loss of base J in WT T . brucei parasites has no effect on siRNA levels [35] , the level of siRNAs derived from the cSSRs significantly increased in the H3 . V KO ( Fig 1A ) . To confirm these changes , we repeated the small RNA-seq analysis of WT and H3 . V KO in triplicate . Quantitation of the small RNAs associated with cSSRs genome-wide indicates a statistically significant increase in 21-27nt RNAs at 30 out of 72 cSSRs in the H3 . V KO ( Fig 1B and S2 Table ) . This can be visualized by specifically mapping the siRNA size range of small RNAs ( S1A Fig ) . The increase in siRNAs is not restricted to cSSRs , but occurs at all previously characterized siRNA generating regions of the T . brucei genome [61 , 62]; including the SLACS and ingi retrotransposable elements , CIR147 centromeric repeats , and inverted repeats ( S2 Fig ) . These regions are also enriched with H3 . V [14 , 68] . Small RNA-seq suggests that the siRNA mapping to dual strand transcription regions at cSSRs is due to continued RNAP II transcription of the convergent PTU resulting in overlapping transcription . An example that illustrates this is shown in Fig 2A–2D , where H3 . V ( and J ) is found enriched at cSSR 2 . 5 . Presumably , H3 . V interferes with RNAP II elongation in these dual strand transcription regions . Removal of H3 . V would then lead to increased nascent RNA corresponding to these regions ( siRNA precursor ) that is then processed , resulting in increased siRNA levels that map to both strands ( Figs 1 and 2D , S1A and S2 Table ) . We have previously utilized strand-specific RT-PCR to follow read-through transcription and increased nascent RNA production in L . major and T . brucei [35] . Here , primers were designed that span the poly ( A ) site of the final gene in the PTU , allowing us to specifically detect nascent , unprocessed RNA within cSSR 2 . 5 ( Fig 2E ) . Strand-specific RT-PCR results indicate an increase in nascent RNA in the H3 . V KO at dual strand transcribed loci at cSSR 2 . 5 as well as cSSR 1 . 4 ( Fig 2F and 2G ) . Although no significant increase in siRNAs was observed in DMOG treated WT T . brucei [35] , we do detect a slight increase in nascent RNA following the loss of base J at cSSR 2 . 5 ( Fig 2F and 2G ) . Nascent RNA is further increased following the loss of both H3 . V and J at the two cSSRs analyzed . Overall these results indicate that H3 . V , and to a lesser extent J , attenuate transcription elongation within dual strand transcribed loci at cSSRs , potentially enabling regulated expression of siRNAs derived from these sites . Loss of H3 . V does not lead to read-through transcription downstream of the dual strand transcribed loci however . We have recently shown that base J is present along with H3 . V at termination sites within a PTU where J loss results in read-through transcription and increased expression of downstream genes in T . brucei [35] . Because RNAP II elongation and gene expression is inhibited prior to the end of these PTUs , we refer to this as PTU internal termination . To explore the impact of H3 . V on RNAP II elongation at these PTU internal termination sites we analyzed the downstream genes by RT-qPCR in the H3 . V KO cell line . At three representative PTU internal termination sites we detect increased expression of downstream genes in the H3 . V KO ( Fig 3A–3D and S3A–S3D Fig ) . In each of these cases gene derepression is enhanced upon the subsequent loss of J in the H3 . V KO following DMOG treatment ( Fig 3D and S3D Fig ) . Importantly , derepression is limited to genes downstream or within the peak of H3 . V/base J . Strand specific RT-PCR using oligos flanking the termination site ( based on mRNA-seq and base J localization [13 , 35] ) detects increased RNA in the H3 . V KO ( Fig 3E–3F and S3F Fig ) . This is consistent with an increase in nascent read-through RNA resulting from continued transcription elongation at PTU internal termination sites marked by base J and H3 . V . Consistent with the gene expression changes , read-through is visibly enhanced at region 7 . 3 upon the subsequent loss of J in the H3 . V KO following DMOG treatment ( Fig 3F ) . These results , combined with our previous study of J regulation of termination [35] , suggest J and H3 . V have independent and overlapping roles in regulating RNAP II termination through the inhibition of transcription elongation and the expression of downstream genes in T . brucei . To further explore the role of H3 . V in the regulation of termination and whether H3 . V functions similarly across the genome , we performed mRNA-seq to compare the expression profiles of WT , WT+DMOG , H3 . V KO and H3 . V KO+DMOG cells ( GEO accession number GSE69929 ) . This led to the detection of 153 mRNAs that are increased at least 2-fold in one or more of the treatments ( S1 Table ) . Many of the gene expression changes have been confirmed by RT-qPCR ( S4 Fig and below ) . Consistent with our previous mRNA-seq results , in WT cells treated with DMOG we observe similar increases in the expression of genes downstream of base J ( and H3 . V ) , which we previously demonstrated is caused by an RNAP II transcription termination defect within a PTU [35] . However , we now see that a significant number of genes downstream of J/H3 . V within other PTUs are upregulated following the loss of H3 . V , and that many are further increased following the subsequent loss of J ( Fig 4 , S1 and S4 Tables ) . In the H3 . V KO we identified 71 genes that are upregulated ( Fig 5A and S1 Table ) . Although many of these genes are not increased by 2-fold in the WT+DMOG condition , some respond at least slightly to the loss of J in WT cells: 28 of the 71 genes upregulated in the H3 . V KO are also increased at least 1 . 3-fold in the WT+DMOG condition , suggesting J and H3 . V have overlapping functions in regulating termination at these sites , where H3 . V plays a dominant role . Consistent with this , 42 of the 71 H3 . V regulated genes are upregulated even further upon subsequent loss of base J in the H3 . V KO . This trend is evident in the heatmap shown in Fig 5A . A specific example is shown in Fig 5B–5F where a cluster of genes is affected by the loss of H3 . V , and to a lesser extent base J . Both chromatin marks are enriched upstream of and within this gene cluster , which consists of genes annotated as VSG pseudogenes , an atypical VSG , and a hypothetical protein . mRNA-seq indicates gene upregulation is largest in the absence of H3 . V and J , which we confirmed by RT-qPCR ( Fig 5E and 5F ) . An additional example is shown in S5 Fig Chromosome maps indicating the genomic location of upregulated genes upon the loss of H3 . V and J are shown in S6 Fig . These results also confirm our initial analyses of nascent and steady-state RNA indicating a role for H3 . V ( and J ) regulating termination and expression of downstream genes ( Fig 3 and S3 Fig ) . 80% of the affected genes are found within 10 kb of H3 . V and base J ( see S1 Table and S7 Fig ) and thus fit a model where derepression occurs as a result of deregulated transcription elongation/termination within a PTU following the loss of the two chromatin marks . In support of this model , strand-specific RT-PCR analysis links gene derepression with increased nascent RNA production downstream of J/H3 . V marks within the PTU ( Figs 3E–3F and S3F ) . In a few cases , genes that are further than 10 kb downstream of base J and H3 . V are still within a regulated cluster , explaining why some genes are indicated as not adjacent to J/H3 . V but can still be regulated by these epigenetic marks ( e . g . genes Tb427 . 07 . 6730-Tb427 . 07 . 6780 , S1 and S4 Tables ) . Although clusters of genes downstream of a J and H3 . V enriched region are similarly upregulated in many cases , the 153 upregulated genes localize to 91 different PTUs ( S4 Table ) . Therefore , H3 . V and J repress gene expression genome-wide , often repressing the expression of a single gene , usually the last gene , within a PTU . Overall , the data indicate that base J and H3 . V have independent but overlapping roles in regulating RNAP II transcription termination within specific PTUs and enable regulated expression of downstream genes . Although the majority of the differentially expressed genes are upregulated , we also see some downregulated ( S1 Table ) . 35 genes are downregulated by at least 2-fold in the H3 . V KO cells . Unlike the upregulated genes , where many were increased furthest following the combined loss of J and H3 . V , only 7 of the 35 downregulated genes decreased more in the H3 . V KO+DMOG condition . However , 33 of the 35 downregulated genes are found proximal ( within 10kb ) to H3 . V and J . The effect is also locus specific and mainly restricted to arrays of genes transcribed from the same strand . These points indicate a strong link between H3 . V/J localization and gene expression changes observed . A few examples are indicated in S8 Fig . Among the genes with most significant downregulation upon loss of H3 . V were procyclins and procyclin associated genes ( PAGs ) , which include surface protein encoding genes most highly expressed during the ( procyclic ) insect stage of the parasite [69] . Although PAGs are located in multiple copies in the genome , within RNAP II transcribed arrays or within RNAP I transcribed procyclin arrays [69 , 70] , the PAGs downregulated following the loss of H3 . V ( PAG1 , PAG2 , PAG4 and PAG5 ) are specifically arranged in an RNAP I transcribed array . For example , there are two PAG2 genes located on chromosome 10 , one in an RNAP II transcribed PTU and the other in the RNAP I transcribed procyclin locus . The only gene that is significantly downregulated when H3 . V is deleted is the one within the RNAP I procyclin locus . A similar locus specific alteration of PAG expression was seen upon the depletion of histone H1 [71] . Interestingly , the PAGs within this locus undergo overlapping RNAP II and RNAP I transcription , i . e . continued RNAP II transcription of the upstream opposing PTU produces antisense PAG RNAs [72] . This suggests a possible mechanism of PAG ( and procyclin ) downregulation resulting from increased formation of dsRNAs upon the loss of H3 . V ( see discussion ) . Similarly , other downregulated genes are arranged in opposing transcriptional genes pairs where extended RNAP II transcription would result in dsRNA for each mRNA ( S8 Fig ) . We also identified 25 genes that are downregulated specifically in the H3 . V KO+DMOG condition , though 22 of these genes are not located near H3 . V or J . We therefore assume many of these changes are an indirect effect of genes that are upregulated in this cell line . For example , we have demonstrated that the genome-wide increase in RNAP II transcription in T . cruzi results in a global increase in gene expression that includes proteins that degrade specific mRNAs [24] . In addition to RNAP II termination sites , H3 . V co-localizes with base J at telomeres , which in the T . brucei genome contain the RNAP I transcribed polycistronic units involved in antigenic variation ( so called VSG expression sites , ESs ) ( Fig 6A ) [27 , 28] . mRNA-seq results indicate that the deletion of H3 . V leads to increased expression of VSG genes from silent ESs , which we have confirmed by RT-qPCR ( S1 Table and Fig 6B ) . Although VSGs within expression sites are not affected by the loss of J in WT cells , five of the VSGs are further upregulated upon the loss of J in the H3 . V KO ( S1 Table and Fig 6B ) . mRNA-seq indicates several expression site associated genes ( ESAGs ) within silent ESs with differential expression . Because the repetitive nature of ESAGs complicates read alignment of mRNA-seq data , we analyzed several genes by RT-qPCR using primers that are specific to ES ESAGs ( S9 Fig ) . Consistent with mRNA-seq results , with the apparent exception of ESAG8 , most of the ESAGs are not upregulated in the H3 . V KO ( S9 Fig ) . The lack of significant change in the majority of ESAGs within the ESs suggests derepression of telomere-proximal VSG genes after H3 . V depletion is not due to transcriptional activation of silent promoters . Repression of silent ESs is mediated in part by the inhibition of RNAP I elongation within the ES preventing the production of VSG mRNA from the silent ESs [73] . Similar to its inhibition of RNAP II transcription elongation at termination sites within PTUs genome-wide , H3 . V may function at telomeric regions to attenuate transcription elongation within the silent ESs , thereby preventing transcription of silent VSGs . To further investigate this possibility and determine how telomere-proximal and telomere distal genes are affected after loss of H3 . V , we used RT-qPCR to compare the derepression of a unique gene in the silent ES15 that is located 10 kb ( VSG pseudogene 11 , Tb427 . BES126 . 13 ) and 1 kb ( VSG11 ) upstream of the telomere ( Fig 6; [74] ) . Although VSG11 was upregulated in the H3 . V KO , we found that the VSG pseudogene is not significantly derepressed in the H3 . V KO or upon the loss of H3 . V and J ( Fig 6B ) . These results overall suggest H3 . V is involved in the repression of VSGs in silent ESs , but does not significantly impact silencing of the entire ES , similar to the role of telomere localized TbRAP1 [75] . ESAG8 derepression is consistent with ES derepression from the promoter , however upregulation of multiple ESAGs would be expected if derepression of the silent ESs occurs from the promoter , which was observed after ISWI depletion [76] . Although we cannot rule out possible alterations in VSG switching in the H3 . V KO ( see discussion ) , the data here suggest H3 . V is involved in repressing RNAP I transcription of VSGs within the silent ESs as well as RNAP II transcription of VSG genes within genome internal PTUs .
H3 . V is a kinetoplastid-specific H3 variant and appears to be the only H3 variant found in these early-diverged eukaryotes . The T . brucei H3 . V shares only 45% sequence identity with the canonical H3 [60] . Although H3 . V localizes to centromeres , it is not essential for viability and does not contain sequence variations common to all identified centromeric H3 variants [14 , 60 , 68] . Aside from its localization to RNAP II termination sites and telomeres , the functional significance of H3 . V and its potential role in the regulation of RNAP II termination has been unexplored . We have demonstrated that H3 . V negatively regulates transcription elongation and promotes RNAP II termination in T . brucei . Several lines of evidence support this conclusion . At many of the same sites within gene clusters where we have previously shown J regulates transcription termination and expression of downstream genes , we detect similar increases in the expression of genes downstream of the termination site following the loss of H3 . V . Also similar to J loss , detection of increased nascent , unprocessed RNA by strand-specific RT-PCR at these sites supports the conclusion that the loss of H3 . V leads to read-through transcription . Furthermore , the loss of H3 . V from regions where dual strand transcription naturally occurs at cSSRs leads to increased levels of siRNAs , and strand-specific RT-PCR indicates this is due to increased transcription of the cSSR . These findings overall suggest H3 . V imparts a repressive chromatin structure that is refractory to transcription elongation , or potentially recruits other repressive factors or transcription termination factors at RNAP II termination sites . To our knowledge , this is the first example of a histone H3 variant that has been shown to repress the expression of mRNAs and non-coding RNAs by promoting transcription termination . These results extend our previous findings that base J functions to prevent read-through transcription at termination sites within gene clusters in T . brucei , revealing an overlapping role of J and H3 . V in the regulation of transcription termination . However , H3 . V appears to have a broader and in many cases a more dominant role . In this study mRNA-seq indicated 71 genes were upregulated by 2-fold or more following the loss of H3 . V . 39% of those genes were also affected by at least 1 . 3-fold following the loss of J alone , and 59% were further increased upon the subsequent loss of base J in the H3 . V KO . Thus H3 . V and J appear to function similarly , but independently in the regulation of transcription termination . Although we cannot exclude a potential role of H3 . V ( and J ) in the regulation of RNA processing , the increase in both unprocessed ( nascent ) and processed RNAs ( mRNAs and siRNAs ) strongly suggests H3 . V regulates RNA abundance at the level of transcription and that the defects we observe are not simply due to an alteration of RNA processing . We propose a model in which both H3 . V and base J inhibit RNAP II elongation , and therefore stimulate termination at sites within gene clusters ( S10 Fig ) . According to this model , the loss of J within a gene cluster results in read-through transcription and expression of genes that were previously silent , but the presence of H3 . V within the downstream cSSR prevents increased dual strand transcription and thus siRNAs are not significantly increased . Similarly , the loss of H3 . V leads to read-through transcription at termination sites within a gene cluster and subsequent gene derepression . H3 . V loss also results in increased transcription of dual strand transcribed loci at cSSRs , giving rise to more siRNAs . Therefore , base J is a chromatin modification that specifically regulates the expression of a subset of genes in the bloodstream form of T . brucei parasites , whereas H3 . V regulates a similar , but larger subset of genes , in addition to the generation of siRNAs . Recent description of the trypanosome stress response has indicated that the location of a gene within a PTU can impact its expression , presumably via regulated transcription elongation [77] . Here we provide evidence that regulated transcription and expression of genes within PTUs can be achieved through their spatial organization and position relative to H3 . V and base J . However , the biological significance of the gene expression changes we describe following the loss of these chromatin marks , remains unclear . It does not help that the majority of the regulated genes are annotated as hypothetical proteins of unknown function . Many of the H3 . V and base J regulated genes include VSGs , ESAGs , RHS proteins , and pseudogenes that are normally lowly expressed ( or not at all ) in wild type T . brucei . Interestingly , consistent with base J synthesis , VSGs and ESAGs are developmentally regulated , typically exclusively expressed in bloodstream form trypanosomes from the telomeric PTU ( VSG ESs ) . Monoallelic expression of a VSG ES leads to the expression of a single VSG on the surface of the parasite , a key aspect of trypanosome antigenic variation . Therefore the repression of silent VSGs by H3 . V/J allows the parasite to maintain this monoallelic expression . Another important aspect of antigenic variation is the periodic switching of the VSG protein expressed on the surface allowing the parasite to remain a step ahead of the host immune response . DNA recombination ( i . e . gene conversion events ) of silent VSG genes into the active ES is the dominant driver of trypanosome antigenic variation . Transcription of a donor DNA sequence has been shown to increase its use during gene conversion events in human cells [78] . It has also been demonstrated that active transcription in T . brucei stimulates DNA recombination [79] . Therefore , the regulation of transcription of silent VSGs by H3 . V/J , including VSGs in silent telomeric ESs and VSG pseudogenes at the end of genomic internal PTUs , could play a role in gene conversion events . It is well characterized that during late phases of mammalian infection , trypanosomes predominately express mosaic VSGs comprised of multiple VSGs and pseudogenes [80] . These findings thus raise the possibility that regulated transcription of silent VSGs by H3 . V/J , in particular VSG pseudogenes at internal PTUs , contributes to gene conversion events that result in the formation of these mosaic VSGs . If H3 . V and base J are utilized to effect specific gene expression changes , then mechanisms likely exist to overcome the silencing effects of these modifications , i . e . regulated addition and/or removal . Histone chaperones and chromatin remodeling complexes incorporate histone variants at specific chromatin locations . Thus , regulation of ( unidentified ) histone chaperones that incorporate H3 . V could enable regulated gene expression . Chromatin remodeling proteins could also be involved in the removal of H3 . V . The first step of J synthesis consists of thymidine oxidation by JBP1 and 2 , which utilize oxygen and 2-oxoglutarate and require Fe2+ as a cofactor . Changes in oxygen concentrations or metabolic changes could thus impact J synthesis and effect gene expression changes . We previously demonstrated oxygen regulation of JBP1/2 and J synthesis , which led to changes in gene expression and pathogenesis of T . cruzi [24 , 31] . JBP1/2 have been shown to have differential chromatin substrates for de novo J synthesis in vivo [13] . Therefore , regulation of JBP1/2 , or associated factors , could provide differential regulation of J synthesis at specific loci . Reiterative oxidations of thymidine residues by JBP1/2 [29] , similar to TET mediated oxidation of cytosines [81] , may also contribute to regulated J synthesis at specific loci . Loss of H3 . V also led to derepression of VSG genes within the silent telomeric ESs . We hypothesize , similar to its effect on RNAP II termination within PTUs genome-wide , H3 . V localized to telomeric repeats limits basal levels of RNAP I transcription elongation within silent ESs [73] . A similar telomeric VSG derepression effect was observed following the loss of RAP1 in T . brucei [75] . The lack of significant ESAG gene derepression suggests loss of H3 . V does not result in derepression from the promoters of silent ESs , though ESAG8 upregulation is consistent with this possibility . Therefore , we acknowledge that further detailed analysis is required , including the use of tagged silent ESs , to fully understand the role of H3 . V on ES transcription . Because of our inability to effectively measure VSG switching rates in our H3 . V KO cell line , we also cannot exclude the possibility that H3 . V restricts VSG switching , though a recent study has indicated that switching frequency does not appear to change significantly in the H3 . V KO or upon the loss of H3 . V and J ( Schulz , Papavasiliou , and Kim , personal communication ) . While base J has no apparent independent role in telomeric repression or VSG switching , the additional derepression of VSG genes observed in the H3 . V KO upon loss of J suggests the novel modified base can act synergistically with H3 . V in telomeric silencing and antigenic variation . This function is consistent with the distinct localization of base J in the silent ESs , with J density highest close to the telomeres [82] . We also found a specific RNAP I transcribed procyclin gene cluster that was downregulated following the loss of J and even more so by the loss of H3 . V . As mentioned above , this locus has been shown to undergo overlapping RNAP II and RNAP I transcription in the procyclic form T . brucei [72] . In the procyclic form , transcription from the opposite strand was detectable from GU2 to EP1 ( S8 Fig ) . Increased antisense transcription from the RNAP II PTU on the opposite strand led to antisense RNA and co-transcriptional silencing of PAG genes [72] . In contrast , there was little transcription from the opposing strand in bloodstream forms . The presence of H3 . V , and to some extent J , may inhibit this dual strand transcription in bloodstream forms , thus reducing the formation of dsRNA and/or transcriptional interference by RNAP II that could interfere with the expression of the procyclin locus . However , analysis of this locus in procyclic cells has indicated that the loss of Argonaute did not appear to alter the expression of the procyclin genes [83] , thus the role of dsRNAs at this locus , if any , remains unknown . Interestingly , several of the other genes that are downregulated in the H3 . V KO are arranged in opposing transcriptional gene pairs where extended RNAP II transcription would result in dsRNA for each mRNA ( S8 Fig ) . Additional studies are needed to characterize the role of H3 . V in regulating transcription at these loci . As we described for the effect of base J in T . cruzi [24] , it is also possible that some transcripts are downregulated in the H3 . V KO due to secondary effects of derepressed genes , which could include regulatory proteins ( destabilizing specific mRNAs ) . Future studies are necessary to further elucidate the mechanisms by which H3 . V regulates transcription , including what proteins interact with H3 . V , the impact of H3 . V on nucleosome structure/stability , whether H3 . V undergoes any post-translational modifications , and how chromatin structure is affected by the loss of H3 . V . It is not clear how sequence variation in H3 . V confers its specific localization to transcription termination sites or its function . H3 PTMs have not been well characterized in T . brucei , however those identified by mass spectrometry analyses include S1 and K23 acetylation and K4 , K32 , and K76 methylation [84] . In comparison to the canonical H3 , the T . brucei H3 . V differs in that it contains an A1 and R23 , and thus lacks the corresponding acetylation . However , differences in PTMs between H3 . V and H3 have not been investigated . Interestingly , the loss of H3 . V and base J in T . brucei did not result in read-through transcription that extends into the downstream gene cluster encoded on the opposite strand and generation of antisense RNAs . In L . major the loss of J alone led to such read-through transcription at a majority of the cSSRs sites in the genome [35] whereas the loss of H3 . V had no effect [85] . Overall these findings indicate that the function of epigenetic modifications in kinetoplastid parasites is not necessarily conserved . One obvious difference between T . brucei and L . major is the absence of a complete RNAi pathway in L . major [86 , 87] . It would therefore be interesting to investigate the role of H3 . V in regulating siRNAs in a Leishmania species with an intact RNAi pathway . In addition to its role in the regulation of dual strand transcription at cSSRs and the generation of siRNAs , we also find H3 . V regulates other characterized siRNA generating loci , including the SLACS and ingi retrotransposable elements , CIR147 centromeric repeats , and inverted repeats . Aside from the SLACS and ingi derived siRNAs , the function of siRNAs in T . brucei is unclear . Mature SLACS and ingi transcripts are present at low levels in WT T . brucei due to the presence of a functional RNAi pathway [88] . Surprisingly , despite the increase in SLACS and ingi siRNAs following the loss of H3 . V , we also observe a modest increase in SLACS ( Tb427tmp . 211 . 5010 ) and ingi transcripts by mRNA-seq ( S1 Table ) . Although the relative increase in steady-state level of siRNAs is greater than that of the mature mRNA transcript , presumably the increased transcription of these loci in the absence of H3 . V increases both RNA species . Dicer 2 is responsible for the formation of siRNAs derived from cSSRs , and its removal ( and corresponding decrease in siRNAs ) did not have a significant effect on the expression of genes located at cSSRs that coincide with the siRNA peak [61] , suggesting that cSSR derived siRNAs do not regulate mRNA abundance . Consistent with this , at cSSRs where we detect increased siRNAs we do not observe significant decreases in mRNAs from genes that overlap the siRNA peak . Therefore the function , if any , of cSSR derived siRNAs remains unknown . In summary , we have provided evidence for the connection between a histone H3 variant and transcription termination for the first time . These findings highlight the importance of chromatin modifications in the regulation of transcription termination , particularly in early-diverged eukaryotes with unique polycistronic transcription . These findings also have direct implications for a strictly post-transcriptional model of gene expression regulation in kinetoplastids .
WT and H3 . V KO bloodstream form T . brucei 221a cell lines of strain 427 were cultured in HMI-9 medium as described previously [89] . The bloodstream form T . brucei H3 . V KO cell line , generated by deleting both H3 . V alleles by homologous recombination [60] , was provided by George Cross . DMOG treatment of cells was performed by supplementing media with 1mM DMOG for 5 days as described previously [35] . Small RNA-sequencing was performed using two different methods . The analysis of WT , H3 . V KO , and H3 . V KO+DMOG ( Figs 1A , 2D and S2 Fig ) was performed as previously described [35] . Briefly , small RNAs were isolated from T . brucei ( 5x107 cells ) using a Qiagen miRNeasy kit according to the manufacturer’s instructions . The small RNA-seq libraries were prepared using approximately 250ng small RNA by Vertis Biotechnology AG , Germany . The small RNA sample was poly ( A ) -tailed using poly ( A ) polymerase . Then , the 5'PPP and cap structures were removed using tobacco acid pyrophosphatase ( TAP , Epicentre ) . Afterwards , an RNA adapter was ligated to the 5'-monophosphate of the RNA . First-strand cDNA synthesis was performed using an oligo ( dT ) -adapter primer and the M-MLV reverse transcriptase . The resulting cDNAs were PCR-amplified to about 10–20 ng/μL using a high fidelity DNA polymerase . The cDNAs were purified using the Agencourt AMPure XP kit ( Beckman Coulter Genomics ) . Quality and concentration of all libraries was determined by capillary electrophoresis and high throughput sequencing was performed on a HiSeq2000 ( Illumina ) . Sequencing reads were mapped to the T . brucei reference genome using Bowtie2 version 2 . 2 . 3 with local sensitive mode , all other parameters default , [90] and further processed using Samtools 1 . 2 [91] . Reads shorter than 18 bp were discarded before mapping . Genome and gene annotations of strain 427 version 6 . 0 were downloaded from EuPathDB [92] and used as the reference in all small RNA-seq analyses . RPM were calculated using a window size of 101 bp and a step size of 101 bp . Total sequence reads and overall alignment rate for all RNA-seq libraries discussed in this publications are listed in S3 Table . Small RNA-sequencing of the triplicate analysis of WT and H3 . V KO ( Figs 1B , 1C , S1A and S2 Tables ) was performed in a similar manner . Briefly , total RNA was isolated from log phase T . brucei cultures ( 5x107 cells ) using Trizol according to the manufacturer’s instructions . The small RNA-seq libraries were prepared using approximately 250ng total RNA using the Illumina-compatible NEBNext small RNA library preparation kit following the manufacturer protocol ( New England Biolabs ) . Quality and concentration of all libraries was determined using a Bioanalyzer 2100 ( Agilent ) . Libraries were pooled using equi-molar amounts and sequenced on a NextSeq500 ( Illumina ) . Both library construction and sequencing were done at the Georgia Genomics Facility ( GFF ) . Small RNA reads were quality and adapter trimmed using Cutadapt [93] and reads shorter than 18 nucleotides were discarded . Reads were mapped to the T . brucei reference genome using Bowtie2 version 2 . 2 . 3 with the following parameters “-a -D 10 -R 5 -N 1 -L 15 -i S , 1 , 0 . 50”[90] and further processed using Samtools 1 . 2 [91] , BEDTools [94] , and custom scripts . RPM shown in Fig 1B and 1C were calculated by dividing the total number of reads in each size class by the total million reads mapped . For Fig 1B , only reads that mapped to the dual strand transcribed region on cSSR 11 . 9 ( 3826–3835 kb ) were included , whereas Fig 1C includes all mapped reads . Differential expression analysis on small RNA-seq read count data ( WT versus H3 . V KO ) was performed using EdgeR ( S2 Table ) . Significance testing was pairwise using Fisher’s Exact test . Significance was assessed in both the total small RNA-seq reads and in the 21-27nt reads . For the mRNA-seq , total RNA was isolated from log phase T . brucei cultures ( 5x107 cells ) using Trizol . 12 mRNA-seq libraries were constructed ( triplicate WT , WT+DMOG , H3 . V KO , and H3 . V KO+DMOG ) using Illumina TruSeq Stranded RNA LT Kit following the manufacturer’s instructions with limited modifications . The starting quantity of total RNA was adjusted to 1 . 3 μg , and all volumes were reduced to a third of the described quantity . High throughput sequencing was performed at the Georgia Genomics Facility ( GFF ) on a NextSeq500 ( Illumina ) . Raw reads from mRNA-seq were first trimmed using Trimmomatic version 0 . 32 [95] . The single-end reads were trimmed for TruSeq3 adapters; leading and trailing bases with quality less than 15 and reads with average quality less 20 were removed . Finally , any reads shorter than 50 base pair were discarded . Remaining reads were locally aligned to the T . brucei Lister 427 version 9 . 0 genome , from EuPathDB [92] , using Bowtie2 version 2 . 2 . 3 [90] . All settings were default except specifying sensitive local and further processed with Samtools 1 . 2 [91] . Transcript abundances were computed using the Cufflinks suite version 2 . 2 . 1 [96] . For individual replicates , Cuffnorm was used with the library type fr-firststrand flag and the T . brucei Lister 427 version 9 . 0 annotation ( downloaded from EuPathDB [92] ) . To estimate gene expression levels for a condition , replicates were used together and analyzed by Cuffdiff with the T . brucei Lister 427 version 9 . 0 annotation . Default parameters were used except specifying library type fr-firststrand . All p values reported here , determined by Cuffdiff , reflect the FDR-adjusted p value . Correlation coefficients for mRNA-seq replicates of WT , WT+DMOG , and H3 . V KO were all greater than 0 . 96 and H3 . V KO+DMOG replicates were greater than 0 . 91 . To express the transcripts levels for individual mRNA encoding genes as shown in S1 Table , we determined the number of reads per kilobase per million reads ( RPKM ) [97] . Briefly , we counted the number of reads mapped to all annotated transcriptomic features ( e . g . mRNA ) on the same strand ( i . e . sense ) and opposite strand ( i . e . antisense ) . Both the sense and antisense read numbers were normalized by length of the feature ( in kilobase ) and the total number of reads ( in millions ) mapped to non-structural RNAs in the corresponding library ( i . e . number of mappable reads excluding rRNA and tRNA reads ) . mRNA-seq data shown in Figs 2C , 3C , 5D , S3C and S5C are from our previously published dataset [35] and are consistent with mRNA-seq performed in this study ( see S3 Table for the RNA-seq datasets used in each figure ) . Genes were considered adjacent to base J and/or H3 . V if the gene , according to the T . brucei Lister 427 annotation , overlapped within 10 , 000 base pairs upstream or downstream of the modification . All J IP-seq and H3 . V ChIP-seq data shown here are from previously published work [13 , 14] . Fold changes for the heatmaps were computed as ( RPKMvar + pseudocount ) / ( RPKMwt + pseudocount ) , where pseudocount = 0 . 5 . Once all fold changes were computed , any fold change value above five was set equal to five to improve visualization . Total RNA was isolated using the hot phenol method , as described previously [98] . To ensure complete removal of contaminating genomic DNA , purified RNA was treated with Turbo DNase , followed by phenol:chloroform extraction . RNA concentration was determined using a spectrophotometer . Strand specific RT-PCR was performed as previously described [99] . ThermoScript Reverse Transcriptase from Life Technologies was used for cDNA synthesis at 60–65°C . 1–2 μg of RNA were used to make cDNA using a reverse primer as described in the Figure legends . PCR was performed using GoTaq DNA Polymerase from Promega . A minus-RT control was used to ensure no contaminating genomic DNA was amplified . Primer sequences used in the analysis are available upon request . Total RNA was obtained using Qiagen RNeasy kits according to manufacturer’s instructions . First-strand cDNA was synthesized from 1 μg of total RNA using an iScript cDNA synthesis kit ( Bio-Rad Laboratories , Hercules , CA ) per the manufacturer's instructions . Quantification of selected genes were performed on an iCycler with an iQ5 multicolor real-time PCR detection system ( Bio-Rad Laboratories , Hercules , CA ) . Primer sequences used in the analysis are available upon request . The reaction mixture contained 5 pmol forward and reverse primer , 2x iQ SYBR green super mix ( Bio-Rad Laboratories , Hercules , CA ) , and 2 μl of template cDNA . Standard curves were prepared for each gene using 5-fold dilutions of known quantity ( 100 ng/μl ) of WT DNA . The quantities were calculated using iQ5 optical detection system software . | Trypanosoma brucei is an early-diverged parasitic protozoan that causes African sleeping sickness in humans . The genome of T . brucei is organized into polycistronic gene clusters that contain multiple genes that are co-transcribed from a single promoter . Because of this genome arrangement , it is thought that all gene regulation in T . brucei occurs after transcription at the level of RNA ( processing , stability , and translation ) . We have recently described the presence of a modified DNA base J and variant of histone H3 ( H3 . V ) at transcription termination sites within gene clusters where the loss of base J leads to read-through transcription and the expression of downstream genes . We now find that H3 . V also promotes termination prior to the end of gene clusters , thus regulating the transcription of specific genes . Additionally , H3 . V inhibits transcription of siRNA producing loci . Our data suggest H3 . V and base J are utilized for regulating gene expression via terminating transcription within polycistronic gene arrays and regulating the synthesis of siRNAs in trypanosomes . These findings significantly expand our understanding of epigenetic regulatory mechanisms underlying transcription termination in eukaryotes , including divergent organisms that utilize polycistronic transcription , providing the first example of a histone variant that promotes transcription termination . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| []
| 2016 | Histone H3 Variant Regulates RNA Polymerase II Transcription Termination and Dual Strand Transcription of siRNA Loci in Trypanosoma brucei |
Stem cells and their niches constitute units that act cooperatively to achieve adult body homeostasis . How such units form and whether stem cell and niche precursors might be coordinated already during organogenesis are unknown . In fruit flies , primordial germ cells ( PGCs ) , the precursors of germ line stem cells ( GSCs ) , and somatic niche precursors develop within the larval ovary . Together they form the 16–20 GSC units of the adult ovary . We show that ecdysone receptors are required to coordinate the development of niche and GSC precursors . At early third instar , ecdysone receptors repress precocious differentiation of both niches and PGCs . Early repression is required for correct morphogenesis of the ovary and for protecting future GSCs from differentiation . At mid-third instar , ecdysone signaling is required for niche formation . Finally , and concurrent with the initiation of wandering behavior , ecdysone signaling initiates PGC differentiation by allowing the expression of the differentiation gene bag of marbles in PGCs that are not protected by the newly formed niches . All the ovarian functions of ecdysone receptors are mediated through early repression , and late activation , of the ecdysone target gene broad . These results show that , similar to mammals , a brain-gland-gonad axis controls the initiation of oogenesis in insects . They further exemplify how a physiological cue coordinates the formation of a stem cell unit within an organ: it is required for niche establishment and to ensure that precursor cells to adult stem cells remain undifferentiated until the niches can accommodate them . Similar principles might govern the formation of additional stem cell units during organogenesis .
Stem cells and their niches constitute functional units that underlie adult organ homeostasis and regeneration following injury or disease . Despite their great medical importance , little is known about how stem cell units , which originate from precursor cells , form during development . Understanding the relations between stem cell precursors and niche precursors and uncovering the molecular pathways that govern the behavior of these populations are likely to enhance our potential to use stem cells in cell-based therapies . Here we use the developing ovary of the fruit fly Drosophila melanogaster as a model to investigate how the formation of niches is coordinated with the development of their resident stem cells . The Drosophila ovary has been an influential model for understanding the interactions between stem cells and their niches [1] , [2] . Each fly ovary contains 16–20 units called ovarioles . At the anterior of each ovariole lies a niche , which is composed of Terminal Filament ( TF ) and Cap cells ( Figure 1A , B ) . Niche cells produce the ligand Decapentaplegic ( Dpp , a BMP2/4 homologue ) , which acts as a maintenance factor to 2–3 Germ Line Stem Cells ( GSCs ) that are attached to the cap cells [3] , [4] . Dpp signaling within GSCs is required to repress the major differentiation gene bag of marbles ( bam ) [5] , [6] . When GSCs divide , one daughter cell remains at the niche as a GSC . The second daughter , called a cystoblast , is removed from the niche and initiates the differentiation program by up-regulating bam . Germ cell differentiation can be followed by the expression of bamP-GFP , a GFP reporter construct that recapitulates Bam expression ( Figure 1B ) [7] . The cystoblast divides four incomplete divisions to form a 2- , 4- , 8- , and finally a 16-cell cyst . Cyst divisions are coordinated by the fusome , an intracellular organelle that is round in GSCs and extended or branched in germ line cysts ( Figure 1A , B ) [8] , [9] . While much is known about how the GSC unit functions in the adult , how niche precursors and GSC precursors are controlled prior to the formation of the adult GSC unit is less clear . At early larval stages , both gonadal somatic cells ( the precursors of niche cells ) and Primordial Germ Cells ( PGCs , the precursors of GSCs ) proliferate . Somatic proliferation at this stage is required to allow correct morphogenesis of 16–20 niches , while PGC proliferation is required to generate sufficient GSC precursors that could occupy the forming niches [10] . At mid third larval instar ( ML3 ) , TF differentiation initiates ( Figure 1C , D ) [11] . TF specification continues throughout the late larval period , and by the late third larval instar ( LL3 ) , 16–20 TF stacks have formed ( Figure 1E , F ) [11] . Cap cells form at the base of TF stacks at LL3 . Once TF and Cap cells form , PGCs can attach to them via E-Cadherin , to become the adult GSCs [12] . Excess PGCs that are not attached to Cap cells are not maintained , and differentiate to form the first germ line cysts and egg chambers of the female [13] . While differentiating PGCs express bam ( Figure 1F ) , their fusomes are still round ( Figure 1G , arrowheads ) , indicating that they have not divided to form cysts yet . To maintain PGC proliferation throughout larval development , their premature differentiation is actively repressed . Many of the repressors of PGC differentiation are later required for GSC maintenance; the translational repressors Nanos and Pumilio act in a cell-autonomous manner to repress both PGC and GSC differentiation [14]–[16] . In addition , the somatic cells of the ovary express Dpp . Similar to GSCs , Dpp signaling within PGCs is required for their maintenance [13] , [15] , [17] . Whether some aspects of PGC maintenance are unique to the precursor cells has not been established . In addition , since both niche and GSC precursors pass through an initial proliferation stage , followed by differentiation , it is unclear whether , or how , those two stages are coordinated between the two populations of cells . Such coordination is required for correct ratios of niches and GSCs , as well as for the correct maintenance of GSCs and their precursors . In a screen that was designed to find novel regulators of niche and PGC development , we found that target genes of the ecdysone pathway affected PGC maintenance . Ecdysone is a steroid hormone that controls many aspects of larval development , which include temporal control of molting as well as regulating cell fate specification and organ morphogenesis [18] , [19] . Ecdysone production in the prothoracic gland is regulated by the brain-derived neuropeptide Prothoracicotropic Hormone ( PTTH ) [20] . This brain-gland connection is reminiscent of the Hypothalamus-Pituitary link in mammals , which is connected to the gonad in a Hypothalamus-Pituitary-Gonadal ( HPG ) axis . The HPG axis and hormonal regulation play a major role in the initiation of adult reproduction in mammals . No role for the steroid hormone ecdysone has been suggested in the initiation of oogenesis in flies . However , recent reports demonstrated that ecdysone signaling is required cell autonomously within adult GSCs for their maintenance and non-cell-autonomously within Escort Cells ( the somatic cells that contact early germ line cysts , Figure 1A ) for correct differentiation of adult GSC daughter cells [21] , [22] . We demonstrate that in the fly , a brain-gland-gonad axis exists , and that ecdysone receptors regulate GSC and niche formation . In the first , proliferative , stage of gonadogenesis , ecdysone receptors are required to repress precocious PGC and niche precursor cell differentiation . Later , ecdysone signaling is required for niche differentiation . Finally somatic ecdysone signaling is required to initiate fly oogenesis in a non-autonomous manner . Combined , ecdysone receptors orchestrate the entire sequence of the formation of the GSC unit in the ovary . Other stem cell units might similarly be organized during development .
To uncover molecular events that underlie niche formation , PGC maintenance , or their coordination , we performed an over-expression screen in larval ovaries ( Supporting Information ) . The driver line traffic jam-Gal4 ( tj-Gal4 ) , which is expressed in the somatic cells of the ovary , but not in PGCs ( Figure 1G ) , was used to generate non-autonomous effects in PGCs . Such effects require large populations of affected somatic cells and might have been undetected by clonal analysis screens . Over-expression of two nuclear hormone receptors , Eip75B ( Figure 1H ) and to a lesser extent Ftz-f1 ( unpublished data ) , in the somatic cells of the ovary resulted in precocious PGC differentiation . In contrast to wild-type ovaries , which contain spherical fusomes ( Figure 1G , arrowheads ) , LL3 ovaries over-expressing Eip75B contained branched fusomes , indicating that PGCs differentiated precociously into germ line cysts ( Figure 1H , arrowheads ) . Eip75B and Ftz-f1 are target genes in the ecdysone response cascade , which times various events throughout embryonic , larval , and pupal life [18] , [19] . This cascade initiates when the hormone ecdysone binds to two nuclear receptors: Ecdysone Receptor ( EcR ) and Ultraspiracle ( Usp ) . Following activation of the EcR/Usp heterodimer , a gene expression program is initiated . Many of the central target genes of this cascade ( including ftz-F1 , Eip75B and broad ) encode transcription factors or nuclear receptors and are common to many tissues . The tissue-specific targets of this signaling pathway are not well characterized . To test whether precocious PGC differentiation resulted from a change in ecdysone signaling , RNAi constructs against EcR or Usp were expressed using tj-Gal4 . The ovary-specific expression ( henceforth termed “somatic expression” ) did not change the timing of the various molting stages , pupation , and hatching . However , extensive differentiation of PGCs was observed in gonads of EcR or usp RNAi animals ( Figure 1I , J , arrowheads ) . While only 2% of control tj>lacZ ovaries contained branched fusomes ( N = 37 ) , 100% of either tj>EcR or usp RNAi ovaries harbored germ line cysts with branched fusomes ( N = 77 and N = 17 , respectively ) . Somatic expression of different RNAi lines against EcR and usp all resulted in PGC differentiation ( Experimental Procedures ) . Recently , ecdysone signaling was shown to maintain adult GSCs in a cell-autonomous manner . To test whether EcR and Usp or their target genes might repress PGC differentiation cell-autonomously , we removed ecdysone signaling components specifically from PGCs . No precocious PGC differentiation was observed when RNAi constructs against EcR and usp , or a dominant-negative isoform of EcRA ( EcRA . W650A , [23] ) , were expressed using the germ-line-specific driver nos-Gal4 . Nor was PGC differentiation observed in PGCs mutant for usp , Eip75B , Eip74EF , or ftz-f1 ( Figure S1 ) . Broad-mutant ovaries also lacked germ line cysts ( see below ) . Thus , during larval stages , ecdysone receptors in the somatic cells of the ovary are required non-autonomously to repress precocious PGC differentiation . In addition to precocious PGC differentiation , precocious niche differentiation also occurred in EcR and usp RNAi ovaries . In wild-type ML3 ovaries , only few cells express the TF markers hedgehog-LacZ ( hh-lacZ ) and Engrailed ( En ) . These cells are still unorganized , and very few short filaments can be detected at this stage ( Figure 2A , B , Table 1 ) . In contrast , removal of EcR or usp from the somatic cells of the ovary by RNAi resulted in more TF cells , which were already organized into filaments by ML3 ( Figure 2C , D , arrows , Table 1 ) . To test whether all aspects of niche formation were precocious , we examined Cap cells , which appear at the larval-pupal transition stage at the posterior base of TFs [24] . Cap cells contain nuclei that are rounder than TF nuclei and also stain with hh-lacZ ( Figure 2E , arrowheads ) . These cells also stain with anti-Tj antibody , which at LL3 stains the Intermingled Cells ( ICs , the cells that directly contact PGCs [10] , [25] ) , indicating that cap cells may originate from anterior ICs ( Figure 2F , inset , arrowheads ) . In EcR ( Figure 2G , arrowheads ) and usp RNAi ( unpublished data ) ovaries , cells with cap cell morphology , which were labeled by hh-LacZ , appeared at the base of precocious TFs already at ML3 . Thus , the development of the entire stem cell niche is precocious when either EcR or Usp are removed from the somatic cells of the ovary . Despite the precocious formation of cap cells in EcR and Usp-RNAi ovaries , we could not observe extra cap cells during larval stages , as has recently been proposed [22] . However , it is possible that increased ecdysone signaling affects cap cell number during pupal or adult stages ( Figure S2 , Text S1 ) . Precocious niche development resulted in disorganization of the anterior part of the ovary . In the wild type , niches are formed as well organized TF stacks , which are regularly spaced throughout the anterior part of the LL3 ovary ( Figure 2H ) . In EcR ( Figure 2I ) or usp RNAi ovaries ( unpublished data ) , TF stacks formed , but some stacks were not positioned correctly from anterior to posterior . In addition , less non-TF cells were present between stacks and anterior to them ( Figure 2I ) . Since TF and cap cells are post-mitotic , we suggest that their precocious differentiation at the expense of the proliferating precursors caused the reduction in anterior size and resulted in morphogenesis defects . Despite their spatial disorder , niches had all their cellular components; we therefore tested whether the precocious niches in EcR and usp RNAi ovaries were functional . Wild type niches secrete Dpp , which results in phosphorylation of Mothers Against Dpp ( pMAD , a SMAD homologue ) within germ cells that are attached to them . We used immunofluorescence labeling to compare the level of pMAD in PGCs that were close to forming niches in wild type and in EcR-RNAi LL3 ovaries . In accord with the normal , albeit early , sequence of niche development , similar levels of pMad were observed in both cases in anterior PGCs that were close to niches ( Table 2 ) . Indeed , in EcR and usp RNAi ovaries , precocious PGC differentiation occurred only in posterior PGCs located away from the niches ( Figure 1I , J ) . Taken together , these data show that removing ecdysone receptors from the somatic cells of the ovary leads to precocious differentiation of both niches and PGCs . Forming niches are functional and protect PGCs that attach to them from differentiation . However , the organization of the anterior of the ovary is defective due to precocious precursor differentiation . To understand how ecdysone receptors repress precocious niche formation and PGC differentiation , we examined the expression of ecdysone receptors and of the transcription factor Broad , an important target of the pathway . Antibodies directed against EcR-A weakly stained all somatic nuclei in mid and late third instar . EcR-B1 was detected in all somatic nuclei during third instar . As expected , no EcR staining was observed within PGCs ( Figure S3 ) . This finding is in accord with the somatic expression of Usp in larval ovaries [26] . The broad locus encodes four different transcripts: broad-Z1 , Z2 , Z3 , and Z4 [27] . An antibody directed against the common region of all Broad isoforms exclusively stained somatic cell nuclei . Staining levels increased as ovaries matured ( Figure 3A , B , C ) . One reason for this increase might be the difference observed in the expression of Broad-Z1 . Staining with anti-Br-Z1 revealed that this isoform was not expressed until ML3 . At ML3 very faint Br-Z1 staining could be observed ( Figure 3D ) , and by LL3 it was strongly expressed in all somatic nuclei ( Figure 3E ) . In contrast , Br-Z1 expression was clearly detected already at ML3 in EcR or usp RNAi ovaries ( Figure 3F , G ) , suggesting that ecdysone receptors repress early expression of Br-Z1 . Significantly , precocious expression was particularly noted in the ICs ( Figure 3F , G , arrowheads ) , which contact PGCs [10] , [25] . Inhibition of Br-Z1 expression by EcR was previously observed in imaginal discs [28] . It was suggested that , in analogy to several mammalian nuclear hormone receptors , EcR and Usp have a dual role: in the absence of ecdysone or when associated with co-repressors , these receptors function as repressors of ecdysone target genes , while in the presence of ecdysone or specific co-activators they promote or have a permissive role in target gene activation [28]–[30] . To test this hypothesis , we used a dominant negative isoform of EcR-A , which cannot bind ligand , and serves as a constant repressor [23] . Indeed , Br-Z1 was not expressed , or expressed in very few cells , in LL3 ovaries expressing the dominant negative EcRA . W650A ( compare Figure 3H to 3E ) . These results demonstrate that EcR and Usp act as early repressors of Br-Z1 and that ecdysone signaling is later required for Br-Z1 expression . Anti-Br-C staining was still observed in EcRA . W650A ovaries ( Figure 3I ) , suggesting that Broad Complex is affected by , but not entirely dependent on , ecdysone signaling [31] . Our results suggest that at early third instar , EcR/Usp mediated repression of Br-Z1 expression delays niche and PGC differentiation , while at late third instar , activation of the ecdysone pathway may promote these events by allowing Broad-Z1 expression . To test this hypothesis and determine the role of active ecdysone signaling and Broad expression in the ovary , we expressed the dominant negative form of each of the three EcR isoforms in the somatic cells of the ovary . The dominant negative form EcRA . W650A produced the strongest phenotypes ( Figure 4A , Figure S4 ) . EcRA . W650A ovaries were markedly smaller as compared to wild type ( 100% of the ovaries , N = 50 , Figure 4A , compare to Figure 1G ) . Very few TF cells , which were not organized into long stacks , were observed in these ovaries ( Figure 4B ) . It has been previously shown that Notch activation is required for cap cell formation [24] , [32] . Indeed , expression of the intracellular portion of Notch in somatic cells markedly increased the number of cap cells forming in wild-type ovaries at LL3 ( Figure 4C , arrowheads , N = 21 ) . However , cap cells were not induced by Notch activation in EcRA . W650A ovaries ( Figure 4D , N = 11 ) , suggesting that somatic ecdysone signaling is required to allow Notch-mediated cap cell formation . The absence of niches in EcRA . W650A ovaries could result from a general developmental arrest , or from a particular problem in niche formation . We therefore tested whether some aspects of gonad morphogenesis did occur properly in EcRA . W650A ovaries . In wild-type ovaries , all somatic cells express the protein Traffic Jam ( Tj ) until ML3 . At this stage the expression of Tj is being limited to ICs [10] , [25] . By LL3 , only ICs , which intermingle with germ cells , express Tj at high levels ( Figure 4E , arrowheads ) . In EcRA . W650A LL3 ovaries , we found Tj-positive cells in the vicinity of PGCs . These cells failed to intermingle with germ cells ( Figure 4F , arrowheads ) . Significantly , the anterior of the ovary was devoid of Tj protein at this stage , indicating that clearance of Tj from the anterior occurred normally . The fact that not all aspects of ovarian maturation were arrested in EcRA . W650A suggests that ecdysone signaling has a more specific role in niche formation . Indeed , mosaic analysis revealed that less TFs formed in ovaries bearing large mutant clones of Eip75B and Ftz-f1 , despite an otherwise normal ovarian development ( Table 1 ) . Their smaller size prompted us to investigate cell proliferation and cell death in EcRA . W650A ovaries . No cell death above wild-type background could be observed ( only 0–4 dying cells are observed in WT , br-RNAi , or EcRA . W650A ML3 ovaries; N = 30 , 26 , and 42 ovaries , respectively ) . However , cell proliferation was significantly reduced in br-RNAi and EcRA . W650A ML3 ovaries . In LacZ control ovaries , an average of 17 . 11 cells were labeled with anti-phospho Histone H3 , a mitotic marker ( SD = 6 . 99 , N = 35 ) . Significantly less cells were in mitosis in EcRA . W650A ( average of 8 . 8 cells , SD = 3 . 47 , N = 35 , t test p = 2 . 53E-8 ) or br-RNAi ( average 14 . 2 cells , SD = 4 . 12 , N = 50 , t test p = 0 . 017 ) . To distinguish between a primary requirement for ecdysone signaling in cell proliferation or in cell differentiation , we forced somatic cells of EcRA . W650A ovaries to proliferate by over-expression of the Insulin receptor ( InR ) . Wild-type ovaries over-expressing InR are larger in size , but their niches are normally patterned ( Figure 4G , arrows ) . In EcRA . W650A ovaries that also expressed InR , ovarian size was similar to that of wild type , indicating that Insulin signaling can overcome the proliferation defect arising from disrupted ecdysone signaling . However , similar to EcRA . W650A , very few TF cells were observed , which were not organized in filaments ( compare Figure 4H arrows to 4B and to wild type , Figure 2H ) . Together with the advanced formation of niches in EcR and usp RNAi ovaries , these results indicate that ecdysone signaling is required for differentiation of somatic niche cells . In addition , ecdysone signaling may also contribute to somatic cell proliferation [33] . As expected , br-RNAi phenotypes were similar to EcRA . W650A in nature but were weaker . br -RNAi ovaries were smaller than wild type and had no TFs , or shorter TFs than wild type ( 100% of the ovaries , N = 25 , compare Figure 4I to Figure 2H ) . Similar phenotypes were observed in ovaries from br1 ( Figure 4J ) or br5 ( unpublished data ) mutant animals , in which Br-Z2 function is removed ( 100% of ovaries , N = 28 for br1 and N = 35 for br5 ) . Importantly , precocious niche formation and PGC differentiation could not be observed in EcR RNAi ovaries that also lacked broad . PGCs in such ovaries contained spherical fusomes and TFs were shorter than wild type ( Figure 4K , L ) , suggesting that Broad is an essential component in ecdysone-mediated control of ovarian morphogenesis . Our results suggest that removal of broad leads to retarded ovarian morphogenesis , while its precocious expression in EcR or usp RNAi ovaries might lead to advanced morphogenesis and to PGC differentiation . To test this directly we over-expressed each of the Broad isoforms in the somatic cells of the ovary . Niche cells were labeled by anti-Engrailed ( En ) and PGC differentiation was monitored using the reporter bamP-GFP [7] . Over-expression of all Broad isoforms led to precocious bamP-GFP expression at ML3 ( 100% of the ovaries , N = 20 , 29 , 30 , and 29 for Br-Z1 , Z2 , Z3 , and Z4 , respectively; compare Figure 5A to 5B for Br-Z1 , 5C for Br-Z4 . Br-Z2 , Br-Z3 not shown ) . Since PGC differentiation was so robust in Broad over-expressing ovaries , we tested the extent to which it could reach . In wild-type adult germaria , Orb is expressed in 8- and 16-cell cysts . When one cell of the 16 is chosen as an oocyte , Orb localizes in this cell ( Figure 5D , arrowheads ) [34] . As expected , anti-Orb staining of wild type LL3 ovaries revealed no Orb labeling ( Figure 5E ) . However , in Br-Z1 ( Figure 5F ) , Br-Z2 , and Br-Z4 ( unpublished data ) over-expressing ovaries , Orb labeling could clearly be seen . Some cysts already localized Orb into one cell ( Figure 5F , arrowheads ) , indicating that PGC differentiation was advanced and could reach the oocyte determination stage . TFs also formed precociously following Broad over-expression ( compare Figure 5A to Figure 5B , Table 1 ) . Interestingly , while Br-Z1 , Z2 , and Z3 expression resulted in both precocious TF and PGC differentiation , Br-Z4 over-expression caused only PGC differentiation , but no change in TFs ( compare Figure 5A to 5C , Table 1 ) . These results further implicate Broad as a major effector of ovarian morphogenesis , and in particular of niche formation and PGC differentiation . To define how somatic ecdysone signaling might induce PGC differentiation , we analyzed its effects on the major germ cell maintenance/differentiation pathway . Similar to GSC maintenance , all PGCs at early larval stages are maintained by Dpp signaling [3] , [15] , [17] , which results in pMad translocation to the nucleus , where it represses bam [5] , [6] . By LL3 , only PGCs that reside at the niche accumulate pMad in their nuclei ( Figure 6A , A′ arrowheads ) . In PGCs that are away from the niche , only background levels of pMad are observed . These PGCs up-regulate BamP-GFP ( Figure 6B ) [7] , [13] . Similar to wild-type ovaries , only a fraction of PGCs in EcRA . W650A LL3 ovaries retained pMad , while most PGCs already down-regulated it . We counted an average of 21 ( SD = 5 , N = 15 ) pMad positive PGCs out of a total of 89 PGCs ( SD = 7 , N = 15 ) . The fraction of pMAD positive PGCs in EcRA . W650A ( 23 . 6% ) is comparable to the percentage of pMAD positive PGCs in wild type ovaries ( 45 pMAD positive PGCs , SD = 8 , N = 11 , which are 22 . 5%–30% out of 150–200 PGCs at LL3 ) . The spatial distribution of pMad labeling was somewhat different in EcRA . W650A ovaries . pMad labeled cells were located mostly next to the few specified TF cells , but some were also detected at the posterior . pMad-positive PGCs were always in contact with somatic cells ( Figure 6C , C′ arrowheads ) . We assume this difference is due to the fact that ICs , which were shown in the adult to mediate Dpp diffusion [35]–[37] , do not intermingle with PGCs in EcRA . W650A ovaries . In addition , pMad levels within PGCs were reduced as compared with wild type PGCs ( Table 2 ) , probably reflecting the reduced amounts of niche cells , which produce Dpp [4] . Strikingly , despite the loss of pMad labeling in 76 . 4% of PGCs , which was comparable to wild type , bamP-GFP was not up-regulated in any of these cells ( Figure 6D ) . Thus , although PGCs lose their major maintenance cue , they delay their differentiation in the absence of somatic ecdysone signaling . This result is particularly intriguing since Mad represses bam transcription directly [5] , [6] . It suggests that PGC maintenance can be uncoupled from PGC differentiation and that other signaling pathways , which are indirectly affected by ecdysone , might integrate on the bam promoter . The dual effect of ecdysone signaling on both somatic cells and PGCs raises the question of how these two processes are connected . One option is that ecdysone signaling , through broad , is only required for somatic niche maturation , which then triggers PGC differentiation . Alternatively , ecdysone signaling and Broad might be required first for niche formation and later , independently , for PGC differentiation . Over-expression of Broad-Z4 resulted in precocious PGC differentiation , without affecting niche formation , suggesting a separate role for ecdysone in the maturation of these two cell populations ( Table 1 , Figure 5C ) . To experimentally test whether PGC differentiation depends on an ecdysone-mediated event that is independent of niche formation , we used a temperature-sensitive Gal80 [38] to temporally control the expression of the dominant negative EcRA . W650A . Larvae were raised in a permissive temperature until niche formation had begun , but before PGCs differentiate ( Materials and Methods , Figure S5 ) . Following a shift to the restrictive temperature , the state of niche development and PGC differentiation was examined . Under these conditions , TFs and cap cells could be observed in both control and experimental ovaries ( Figure 7A , B , arrows ) . These niches were functional , since PGCs that were attached to them maintained pMAD labeling ( Figure 7A , B , arrowheads , N = 36 and N = 25 , respectively ) . In control ovaries , PGCs that were not located close to niches up-regulated bamP-GFP ( Figure 7C , N = 49 ) . However , PGCs in EcRA . W650A ovaries failed to differentiate and did not up-regulate bamP-GFP , despite niche formation ( Figure 7D , N = 56 ) . Similar results were observed with a temperature-sensitive allelic combination of EcR ( EcRA438T/EcRM554fs , unpublished data , N = 25 ovaries ) . These data suggest that PGC differentiation requires wild-type ecdysone signaling even after niches have formed . To understand why PGCs failed to differentiate in EcRA . W650A temperature shifted ovaries , we examined Br-Z1 expression . Br-Z1 was expressed in the anterior of these ovaries and in the formed niches ( Figure 7E , arrows , N = 31 ) . Anterior expression of Br-Z1 in the temperature shift experiments is expected , since in wild-type LL3 ovaries tj-Gal4 expression is weak in these regions ( Figure 1G ) . Significantly , no Br-Z1 could be observed in ICs , which are located posterior to the niches , and where tj-Gal4 is strongly expressed . These results further implicate Br-Z1 expression within ICs , rather than within niches , as required for PGC differentiation at the end of larval development . We used the temperature shift approach to further test the temporal requirement of EcR in gonad morphogenesis and found that somatic expression of EcRAW650A only in the adult resulted in normal ovariole morphology ( Figure S5 ) . Likewise , a defect in ecdysone signaling during larval development could not be corrected by wild type signaling in the adult ovary . Overall , the temperature shift experiments demonstrated an absolute requirement for somatic ecdysone signaling during larval ovarian development . In particular , these experiments demonstrate that ecdysone is required in parallel for niche and PGC differentiation; even when ovarian morphogenesis is normal , and niches do form , an additional ecdysone-mediated event must occur to allow PGC differentiation . The temperature shift experiments suggest that PGCs might differentiate in response to a specific ecdysone pulse , occurring after ML3 and prior to pupation . At least one such pulse has been identified in Drosophila [39] . To test this idea more directly , we timed wild-type PGC differentiation by analyzing the expression of bamP-GFP and found that PGC differentiation coincides with the initiation of wandering behavior . When insect larvae attain a critical body size , an ecdysone pulse triggers distinct behavioral changes that include cessation of feeding and seeking a location for pupation ( wandering behavior ) [40] . 0–4 h prior to the initiation of wandering only 21% of the ovaries contained very few differentiating PGCs ( Figure 7F ) . bamP-GFP levels in these differentiating PGCs were very low , indicating very early stages of differentiation ( Figure S6 ) . In contrast , 0–4 h following the initiation of wandering 85% of larval ovaries contained many differentiating PGCs with strong bamP-GFP labeling ( Figure 7F , Figure S6 ) . The tight temporal correlation between PGC differentiation and wandering behavior suggests that a specific ecdysone peak is required for PGC differentiation and that hormonal regulation is directly involved in initiating oogenesis in flies .
In the forming Drosophila ovary , the ecdysone signaling pathway coordinates somatic niche formation with GSC establishment , leading to the formation of 16–20 stem cell units . The dual function of early repression and late activation of Broad by EcR/Usp allows this pathway to initially repress both niche and stem cell precursors . Later , ecdysone signaling sequentially initiates TF formation and then PGC differentiation ( Figure 8 ) . Within the temporal framework , provided by repeated ecdysone pulses , other signaling pathways may participate in determining the specific rate of precursor cell proliferation and their differentiation . Future work will be needed to determine at what level ecdysone signaling controls these pathways . Our results show that somatic ecdysone signaling elicits a secondary signal that integrates on the major axis of GSC maintenance/differentiation . This signal is required to induce Bam expression in PGCs that are located away from the niche and to initiate their differentiation . Whether ecdysone signaling directly affects the major genes required for niche differentiation remains to be seen . Ecdysone initiates niche formation at ML3 , and PGC differentiation a few hours later . These events do not occur with the earlier peaks of ecdysone , at first and second instar . Gene activation by nuclear hormone receptors is highly context-dependent , and each target gene may require particular co-repressors or co-activators . We hypothesize that the target genes required for the differentiation of niches and PGCs are different , with promoters that require different ligand concentration or different co-activators , which might only be expressed at particular developmental times . Another option ( not mutually exclusive ) is that the target cells for the two roles of ecdysoene ( i . e . , niche formation and PGC differentiation ) are different; clonal analysis suggests that ecdysone signaling is required within TF precursors for their differentiation , while ICs may control PGC differentiation . Several lines of evidence suggest a parallel role of ecdysone in niche and PGC differentiation . First , over-expression of Broad-Z4 leads to PGC differentiation , without affecting niche formation ( Figure 5 , Table 1 ) . Second , our temporal shift experiments demonstrated that niche formation in itself is insufficient to induce PGC differentiation ( Figure 7 ) . Lastly , in EcR and usp-RNAi ovaries , Broad-Z1 is over-expressed mainly in ICs , indicating that this cell population , which is in direct contact with PGCs , is a possible source for a signal inducing PGC differentiation ( Figure 3 ) . What that substance might be is currently under investigation . Activation of the ecdysone signaling pathway in the larva leads to PGC differentiation . In contrast , activation of this pathway in the adult is required to maintain GSCs un-differentiated [21] . Thus , ecdysone signaling serves opposite functions in the adult and in the larva . We have previously demonstrated that many of the mechanisms that maintain GSCs in the adult are already required to maintain PGCs in the larva [15] . Ecdysone signaling is therefore a first regulator that exhibits a reversal of function between a developing stem cell unit and a functional one . The distinct consequence of ecdysone signaling in adult and larval ovaries is reflected in the different manner in which the signal is transmitted . In contrast to the larva , the adult function of ecdysone is cell autonomous and does not seem to strongly rely on Broad function [21] . In addition , somatic ecdysone signaling in the larva is transmitted to PGCs by a signal that integrates downstream of pMad , on the bam promoter ( Figure 6 ) , while in the adult ecdysone signaling affects GSCs upstream of pMad [21] . In addition to a role within GSCs , ecdysone signaling may be required in Escort cells for correct cyst development [22] . Thus , the different physiological conditions during larval development and in the adult lead to very different effects on a forming versus an adult stem cell unit . One other difference between adult and larval ecdysone signaling is the source of ecdysone that reaches the ovary . In the adult , ecdysone is produced locally by developed egg chambers and is affecting GSCs in a physiological positive feedback loop [21] , [41] . In the larva , developed egg chambers do not exist . The temporal correlation of PGC differentiation with the peak of ecdysone that leads to wandering behavior demonstrates that larval ovaries , similar to other larval organs , respond to ecdysone that is produced by the prothoracic gland , located near the fly's brain . This suggests a similarity to vertebrate development that was hitherto unappreciated . In verterbrates , a hypothalamic-pituitary-gonadal axis initiates and accompanies adult reproductive responses [42] , [43] . This work shows that in fruit flies , a brain-gland-gonad axis also operates . The anatomical analogy , however , does not fully extend to the molecular messengers that convey the signals . The hypothalamus-pituitary connection can be equated with the fly neurons that release PTTH into the prothoracic gland , and elicit ecdysone production [44] . Similar to LH and FSH , which are released from the pituitary gland , ecdysone released from the prothoracic glad affects the gonads and is required for the initial differentiation of PGCs ( i . e . , for the initiation of oogenesis ) . Later in adult life , akin to steroid hormones produced by the vertebrate gonad , ecdysone is produced by mature egg chambers [41] . It will be of interest to establish whether the testis in Drosophila males also produces ecdysone . Even prior to the initiation of reproduction in mammals , nuclear receptors are involved in gonadogenesis . Nr5a1 is required for the formation of both the ovary and the adrenal gland [45] , [46] . Interestingly , Nr5a1 is a mammalian homologue of Ftz-f1 , which also has a role in Drosophila gonadogenesis . The physiological role of hormones in niche or stem cell function is not limited to the gonads . Hormones were shown to affect the hematopoietic stem cell niche [47] , and the mammalian homologue of EcR promotes neurogenesis in human embryonic stem cell cultures [48] . Steroid hormones are also required for the regeneration of the mammary gland [49] , [50] . Similar to our results with ecdysone , the effects of hormones on mammary stem cells are probably indirect , through support cells . Whether the analogy could be extended , and these hormones will prove to affect niche development , remains unanswered . Future work will undoubtedly solve this issue , since understanding how niches and stem cells are coordinated by hormones , or other signals , is crucial for the understanding of regeneration and for applicative approaches in cell-based therapies .
tj-Gal4 is a NP insertion line ( P{GawB}NP1624 ) into the traffic jam gene , and was obtained from the Drosophila Genetic Resource Centre . UAS-Broad-Z1 , Z2 , Z3 , and Z4 were generously provided by Dr . Lynn Riddiford ( HHMI , Janelia Farms Research Campus ) . bamP-GFP is a reporter GFP fused to a fragment of the bam promoter . The transgene located on the X chromosome was obtained from Dr . Dennis McKearin . RNAi lines directed against EcR ( 1765R-4 , 1765R-2 ) or Usp ( 4380R-1 ) were obtained from NIG-Fly . RNAi lines against EcR ( 37058 ) , Usp ( 16893 ) , and Broad ( 104648 ) were obtained from VDRC . RNAi line EcR-IR was from Bloomington . Throughout the main text , RNAi lines 1765R-4 , 1765R-2 for EcR , and 4380R-1 for Usp are shown . Somatic expression of EcR-IR and line 16893 resulted in fewer and less developed cysts than lines 37058 , 1765R-2 , 1765R-4 , and 4380R-1 . FRT19A , usp3 was provided by Dr . Oren Schuldiner ( Weizmann Institute ) . FRT80B , Eip74EFDL-1 was provided by Dr . Daniela Drummond-Barbosa ( Johns Hopkins University ) . UAS-Nintra was provided by Dr . Allison Bardin ( Institute Curie ) . br1 , br5 , UAS-EcRA . W650A , UAS-EcRB1 . W650A , UAS-EcRB2 . W650A , Eip75B07041 , and Ftz-f103649 were obtained from the Bloomington Stock Center . UAS-InR and UAS-lacZ were provided by Dr . Jessica Treisman ( NYU School of Medicine ) . Somatic clones were generated using the line c587-Gal4 , UAS-flp ;; FRT2A , ubi-GFP/TM6 . Germ line clones were generated using the line UAS-flp; nos-Gal4; FRT2A , ubi-GFP . usp clones were generated using FRT19A , arm-lacZ; hs-Flp . Clones were induced by heat-shock 48 h AEL , for 30 min at 37°C . To obtain flies in similar developmental stages , care was taken to work with under-crowded cultures . Flies were transferred into a fresh vial to lay eggs for 2 h , and were then removed . Vials were left at 25° for 96 h ( mid third instar , ML3 ) or 120 h ( late third instar , LL3 ) . Under these conditions the development of wild type gonads is uniform . The terminology we use is according to Ashburner [51] and is different from the one used by Zhu and Xie [13] , who go by King [52] . For time course of PGC differentiation , consecutive layings of 2 h were allowed to mature in a 25°C incubator with 70% humidity and 12 h of dark-light cycles . Under these conditions , flies begin wandering behavior at 112 h AEL . For temporal control of EcRA . W650A expression: bamP-GFP;tj-Gal4/UAS-EcRA . W650A;UAS-Gal80ts , flies were cultured for 6 d at 18°C , then shifted to 29°C for an additional day . Alternatively , a regime of 7 d at 18°C , and a shift to 29°C for an additional day was used ( Figure S5 ) . In both cases , larvae were crawling on the bottle walls when dissected . The following monoclonal antibodies were obtained from the Developmental Studies Hybridoma Bank , developed under the auspices of the NICHD and maintained by the University of Iowa , Department of Biology: Monoclonal 1B1 ( developed by Dr . Howard Lipshitz ) antibody is directed against an Adducin ( 1∶20 ) ; LC28 . 26 ( contributed by Dr . Paul Fisher ) anti-LaminC ( 1∶20 ) ; 6H4 ( developed by Dr . Paul Schedl ) anti-Orb antibody ( 1∶20 ) ; 25E9 . D7 anti-Broad Core ( 1∶10 ) , Z1 . 3C11 . OA1 anti-Broad Z1 ( 1∶10 ) developed by Dr . Greg Guild; 15G1a anti-EcRA ( 1∶10 ) , AD4 . 4 anti-EcRB1 ( 1∶10 ) , AG10 . 2 anti-EcRC developed by Drs . Carl Thummel and David Hogness; 4D9 anti-Engrailed ( 1∶20 ) , developed by Dr . Corey Goodman . Rabbit anti-Vasa ( 1∶5000 ) was a gift from Dr . Ruth Lehmann ( HHMI , New York University ) . Rabbit anti-pMAD was a gift from Dr . Ed Laufer ( Columbia University ) . Rabbit anti-β Gal ( 1∶15 , 000 ) was from Cappel . Rabbit anti-GFP ( 1∶1 , 000 ) was from Invitrogen . Secondary antibodies were from Jackson Immunoresearch or from Invitrogen . Unless otherwise specified , all incubations were at room temperature . Ovaries were dissected in Drosophila Ringers Buffer and fixed for 20 min with 5% formaldehyde . Ovaries were then washed once for 10 min with PBS containing 1% Triton-X-100 ( 1% PBT ) , and washed again with 1% PBT for an additional hour . Ovaries were blocked with PBS containing 0 . 3% Triton-X-100 and 1% BSA ( 0 . 3% PBTB ) for 1 h , and then incubated with first antibody in PBTB overnight at 4°C . Ovaries were washed twice in 0 . 3% PBTB for 30 min and then blocked with 0 . 3% PBTB supplemented with 5% Normal Donkey Serum ( NDS ) for 1 h . Secondary antibody was diluted in 0 . 3% PBTB supplemented with 5% NDS . Following 2 h incubation with secondary antibody , ovaries were washed three times in 0 . 3% PBT , 30 min each , and mounted with Vectashield ( Vector Laboratories ) . Confocal imaging was with Zeiss LSM 710 on a Zeiss Observer Z1 . For statistical analyses , two-tailed student's t tests were performed . p values are indicated . | Tissue-specific stem cells reside in specialized microenvironments ( niches ) . How the generation of niche cells and resident stem cells is coordinated , and how their correct numerical ratios are regulated , remains poorly understood . Here , we examine the potential mechanisms of this process by analyzing the formation of the fly ovary . Specifically , we uncover a role for ecdysone , which is a steroid hormone renowned for its role in insect molting but that also controls many aspects of larval development in flies and mammals . We find that ecdysone signaling in fly larvae coordinates the development of niche cells relative to their resident germ line stem cells ( GSCs ) . Ecdysone receptors initially serve as repressors of differentiation , allowing precursor cells of both niches and GSCs time to proliferate and attain correct cell numbers . Later , ecdysone receptors allow differentiation of niches while simultaneously maintaining GSC precursors in an undifferentiated state , until the newly formed niches can accommodate them . Finally , ecdysone induces the differentiation of GSC precursors that are not incorporated in niches . Our work exemplifies one possible solution to coordinating stem cell and niche development: using a common signal to affect both cell types . A further understanding of these and other mechanisms will offer novel insights into regeneration and could help guide cell-based therapies for various diseases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"animal",
"models",
"developmental",
"biology",
"drosophila",
"melanogaster",
"model",
"organisms",
"organism",
"development",
"stem",
"cells",
"molecular",
"development",
"genetics",
"biology",
"morphogenesis",
"genetics",
"and",
"genomics",
"cell",
"differentiation",
"gene",
"function"
]
| 2011 | Coordinated Regulation of Niche and Stem Cell Precursors by Hormonal Signaling |
UNC-104/KIF1A is a Kinesin-3 motor that transports synaptic vesicles from the cell body towards the synapse by binding to PI ( 4 , 5 ) P2 through its PH domain . The fate of the motor upon reaching the synapse is not known . We found that wild-type UNC-104 is degraded at synaptic regions through the ubiquitin pathway and is not retrogradely transported back to the cell body . As a possible means to regulate the motor , we tested the effect of cargo binding on UNC-104 levels . The unc-104 ( e1265 ) allele carries a point mutation ( D1497N ) in the PI ( 4 , 5 ) P2 binding pocket of the PH domain , resulting in greatly reduced preferential binding to PI ( 4 , 5 ) P2 in vitro and presence of very few motors on pre-synaptic vesicles in vivo . unc-104 ( e1265 ) animals have poor locomotion irrespective of in vivo PI ( 4 , 5 ) P2 levels due to reduced anterograde transport . Moreover , they show highly reduced levels of UNC-104 in vivo . To confirm that loss of cargo binding specificity reduces motor levels , we isolated two intragenic suppressors with compensatory mutations within the PH domain . These show partial restoration of in vitro preferential PI ( 4 , 5 ) P2 binding and presence of more motors on pre-synaptic vesicles in vivo . These animals show improved locomotion dependent on in vivo PI ( 4 , 5 ) P2 levels , increased anterograde transport , and partial restoration of UNC-104 protein levels in vivo . For further proof , we mutated a conserved residue in one suppressor background . The PH domain in this triple mutant lacked in vitro PI ( 4 , 5 ) P2 binding specificity , and the animals again showed locomotory defects and reduced motor levels . All allelic variants show increased UNC-104 levels upon blocking the ubiquitin pathway . These data show that inability to bind cargo can target motors for degradation . In view of the observed degradation of the motor in synaptic regions , this further suggests that UNC-104 may get degraded at synapses upon release of cargo .
Transport of pre-synaptic vesicles from the neuronal cell body to the synapse is an essential process to ensure that the nerve terminals can effectively participate in synaptic transmission [1] , [2] . This transport is a regulated process that occurs primarily using the Kinesin-3 family motor UNC-104 , Imac , KIF1A and KIF1Bβ , respectively , in the model systems C . elegans , Drosophila , mouse and humans [3]-[9] . In C . elegans , mutants in unc-104 have locomotory defects that arise from the absence of transport of synaptic vesicles , leading to reduced synaptic transmission at neuromuscular junction synapses [3] , [10] . Molecular motors in neurons such as UNC-104 are thought to bind to their cargoes in the cell body of the neuron , get transported along microtubule tracks to synapses and release their cargo upon reaching the synapse [2] . It has been proposed that upon release of cargo the motor gets either inactivated or degraded [11] , thus suggesting cargo binding and cargo release as possible means to regulate motor levels . UNC-104 recognizes its cargo by binding PI ( 4 , 5 ) P2 present on the carrier vesicle via its PH domain [12] and its mammalian orthologue in addition uses other proteins to recognize cargo [13] . Several effects of cargo binding on the Kinesin-3 family motors have been shown . Cargo binding by a chimeric Kinesin-3 leads to aggregation of the motor on the cargo surface and improved processivity of the chimera [14] , [15] . Mutations in the cargo-binding PH domain of UNC-104 that do not bind PI ( 4 , 5 ) P2 efficiently have also been suggested to affect processivity of the motor [12] , [14] . Further , it has been proposed that UNC-104 dimerizes upon cargo binding [14] . The mammalian KIF1A has recently been reported to exist in a dimeric autoinhibited state from which it is released upon cargo binding [16] , [17] , showing that while the orthologues behave differently , they are both regulated by cargo binding . Similarly another motor , Kinesin-1 , is maintained in an inactive folded state [18] and is activated by binding to regulatory molecules/cargo adaptors . Simultaneous binding by both JIP1 and Fez1 activates Kinesin-1 and allows the motor to bind microtubules [19] . Cargo release has also been postulated to play important roles in motor regulation [20] . Motors involved in anterograde axonal transport such as Kinesin-1 , Kinesin-3/KIF1A and heterotrimeric Kinesin are all thought be regulated after releasing cargo at the synapse . All three motors , although transported robustly in the anterograde direction to the synapse , are not efficiently retrogradely transported [6] , [21]-[23] . These observations have led to the hypothesis that once these motors release cargo at synapses , they are largely degraded , thus maintaining directionality of axonal transport [11] . We sought to test this hypothesis for the C . elegans UNC-104 motor protein . To do so , it is necessary to address the following two questions . 1 ) Does the motor get degraded at the synapse ? 2 ) Does the motor get degraded once there is no binding to the cargo ? To answer the first question , we established that the wild type motor is degraded in synaptic regions and that it does not return to the cell body from the synapse . We further showed that the degradation near the synapse takes place through the ubiquitin pathway . To address the second question , we studied the effect of lack of cargo binding on the C . elegans UNC -104 motor protein . For this we used a mutant UNC-104 motor and showed that it has greatly reduced ability to preferentially bind PI ( 4 , 5 ) P2 in vitro as well as greatly reduced presence on pre-synaptic vesicles in vivo . We found that this leads to almost total loss of the motor in vivo , even though the motor still retains the ability to bind other lipids . The relationship between ability to bind cargo and motor levels was verified by analyzing two intragenic suppressors of the original mutation in the PH domain . The suppressors only moderately reduce the ability to preferentially bind PI ( 4 , 5 ) P2 and we see that UNC-104 levels are partially restored . All three PH domain variants of the motor are degraded via the ubiquitin pathway in synapse rich regions of the animal . A triple mutant reversing the effect of one of the suppressor mutations again does not preferentially bind PI ( 4 , 5 ) P2 in vitro , does not provide behavioural rescue and does not show expression of UNC-104 in vivo . These findings , together with the observed degradation of wild type UNC-104 in synaptic regions , suggest that the synaptic vesicle motor UNC-104 is degraded upon release from pre-synaptic vesicles near the synapse .
To determine whether the UNC-104/KIF1A motor is degraded at synapses we used a transgenic line over-expressing UNC-104::GFP in the six mechanosensory neurons of C . elegans . We examined the posterior neurons ( PLM ) whose morphology and synaptic locations are very well defined [24] . Further , a C-terminal UNC-104::GFP fusion provides functional rescue and its localization is similar to that of endogenous UNC-104 [25] suggesting that the addition of GFP does not impair the motor's in vivo function or localization . In a wild type background the UNC-104::GFP is present in the cell body , neuronal process and at synaptic regions ( Figure 1A: b1-b3 ) . To determine whether UNC-104 is degraded we crossed the transgenic strain expressing UNC-104::GFP into a temperature sensitive uba-1 ( it129ts ) mutant . uba-1 encodes the only C . elegans E1 ubiquitin activating enzyme [26] . This activation is an early and essential step in the ubiquitin-degradation pathway . Consequently in uba-1 animals ubiquitin-mediated degradation is reduced . At the lower growth temperature of 16°C the expression of UNC-104::GFP in uba-1 animals is not significantly different from wild type in the cell body , neuronal process or at synaptic regions ( Figure 1A: b1-b6 , 1B ) . However , at the restrictive temperature of 22°C the expression of UNC-104::GFP significantly increases in synaptic regions ( Figure 1A: e1 , e4 , 1B ) . The expression remains largely unchanged in the axon and in the cell body , although our method may not be sensitive to small changes in protein levels , especially in the narrow geometry of the neuronal process ( Figure 1A: e2 , e5 , 1C , Figure S4H ) . To confirm that the morphology of the mechanosensory neuron ( including its synapses ) is relatively unaffected in uba-1 animals , we examined the localization and levels of soluble GFP and of the synaptic vesicle marker GFP::RAB-3 [27]-[29] . No alteration in expression levels of soluble GFP or GFP::RAB-3 was observed in the synapses , cell body or axon in uba-1 animals ( Figure 1A: d1-d6 , f1-f6 , 1C , 1D , Figure S4F , S4G , S4H ) . Compared to wild type , no changes were observed in the area and intensity of GFP in synaptic regions marked either by soluble GFP or by GFP::RAB-3 in uba-1 animals , with the exception of a modest decrease observed in synaptic area marked by GFP::RAB-3 at the restrictive temperature ( Figure 1A: d1-d6 , f1-f6 , 1B , 1C , 1G ) . This exception is consistent with the known importance of degradation for synapse formation in mechanosensory neurons [30] . Taken together these data suggest that development of the mechanosensory neurons and their synapses are not greatly altered in uba-1 ( it129ts ) while there are significant effects on the levels of expression of the synaptic vesicle motor UNC-104 at synaptic regions . The above observations also suggest that UNC-104 may get degraded directly through attachment of ubiquitin ( 8 kDa ) molecules to the motor . To test if UNC-104 is ubiquitinated we immunoprecipitated the endogenous UNC-104 motor ( approximately 200 kDa ) from a mixed-stage C . elegans extract . Western blot analysis of immunoprecipitated UNC-104 motor showed that the same band of about 200 kDa was recognized by both the anti-UNC-104 and the anti-ubiquitin antibodies ( Figure 2D ) . Further , western analysis of the immunoprecipitate obtained using anti-ubiquitin and probed with anti-UNC-104 showed an approximately 200 kDa band , which migrates identically to the endogenous UNC-104 motor ( Figure 2D ) . However , unlike the immunoprecipitation with the anti-UNC-104 , immunoprecipitation using anti-ubiquitin showed the presence of UNC-104 in the supernatant . This signal may rise from UNC-104 molecules that are not ubiquitinated . Our observations suggest that UNC-104 can be ubiquitinated in vivo . Further , the data imply that UNC-104 transports synaptic vesicles to the synapse and upon reaching that location UNC-104 is degraded through the ubiquitin pathway , possibly through direct ubiquitination of the endogenous motor . Our observation that the motor reaching the synapse gets degraded predicts that there would be little retrograde transport of UNC-104 from the synapse back to the cell body . To test this hypothesis we carried out a transport assay by laser microsurgery of the mechanosensory neuron . We had observed that 1 hour after axotomy , cargoes such as GFP::RAB-3 and SNB-1::GFP accumulate on both sides of the cut site [31] . By contrast , wild type UNC-104::GFP accumulates only in the proximal region , i . e . , at the end of the cut that is attached to the cell body ( Figure 2A: b , d arrowhead ) . The distal end shows no accumulation of UNC-104::GFP , corroborating the hypothesis . Further the distal axon shows much lower level of UNC-104::GFP than the same region of the uncut axon ( Figure 2A: b , d arrow ) . This reduction could result from UNC-104 in the distal axon being degraded after reaching the synapse . UNC-104::GFP has been shown in the seconds time-scale to undergo microscopic motion in both anterograde and retrograde directions but with a significant anterograde bias [25] , which may result in overall bulk flow of the motor also being biased towards synaptic regions . Our results show that additionally , degradation of the motor in the synaptic region confers a macroscopic directionality to the movement of the motor . To further test this explanation , we carried out laser axotomy in uba-1 animals and also did bleach recovery experiments to assess UNC-104 motor flow . uba-1 animals , one hour after laser axotomy of mechanosensory neurons expressing UNC-104::GFP , showed robust levels of UNC-104::GFP at the proximal regions ( Figure 2A: f , i arrow ) and significant levels in the distal regions as well ( Figure 2A: e-i arrowhead ) . Thus upon blocking ubiquitination , UNC-104 is capable of macroscopic retrograde movement . To further confirm predicted trends in macroscopic motor movement in an uninjured neuron , we carried out a bleach recovery experiment of UNC-104::GFP in both wild type and uba-1 animals . In either genotype , the UNC-104 motor recovers in both anterograde and retrograde directions in the time frame of seconds ( Figure 2B ) . In wild type , the anterograde recovery is faster than the retrograde recovery , consistent with prior observations of anterograde bias of the microscopic movements of UNC-104::GFP ( Figure 2C ) [25] . Further , supporting our hypothesis , we observed that the retrograde recovery front moved faster in uba-1 animals compared to wild type , while the anterograde recovery in the two genotypes did not differ significantly ( Figure 2B , 2C ) . These observations show that in wild type , very little of the anterograde motor UNC-104 is transported back to the cell body from the synapse . By contrast , when ubiquitin-mediated degradation is blocked , there is significant retrograde transport of the motor from the synapse towards the cell body , likely due to increased UNC-104 levels at synapses . After establishing that the UNC-104 motor is degraded at synapses , we wished to study a possible mechanism for this process . One hypothesis is that once the motor gets to the synapse , it releases cargo and is then targeted for degradation [11] , suggesting that degradation of the motor is linked to its being unbound to cargo . We decided to test this by studying the fate of the UNC-104 motors in a series of alleles that either strongly or moderately alter the ability of the motor to bind cargo through its PH domain . We first attempted to identify a pre-existing allele affecting cargo binding by sequencing several unc-104 alleles ( Figure S1A ) . Of these , unc-104 ( e1265 ) , a canonical allele , showed a single amino acid change D1497 to N in the PH domain ( Figure S1A , S1B ) . To test participation of the highly conserved residue D1497 in binding PI ( 4 , 5 ) P2 , we built a homology model of the UNC-104 PH domain using the crystal structure of the closest orthologue in the database , the protein DAPP1/PHISH ( Figure 3A , Figure S1C ) [32] , [33] . The residues ( KK1463/4 and R1496 ) , known to be important for lipid binding [12] , are respectively , 12 Å/16 Å and 3 . 8 Å from D1497N ( Figure 3A ) . Thus the residue D1497 is on the surface of the PH domain in a region known to be important in binding PI ( 4 , 5 ) P2 . Further , on docking the ligand PI ( 4 , 5 ) P2 on to the homology model using GRAMM [34] we observed that in 40% of the models it preferentially binds to the region juxtaposed to the D1497 , R1496 and KK1463/4 residues ( Figure 3A ) . The next two most common models ( 25% , 20% ) identified for ligand docking do not show proximity to residues known to be important in PH domain-PI ( 4 , 5 ) P2 interactions . To directly test the role of the D1497N mutation encoded by the unc-104 ( e1265 ) allele ( which encodes the protein UNC-104 ( D1497N ) ) in binding to PI ( 4 , 5 ) P2 , we carried out an in vitro liposome binding assay . The wild type UNC-104 PH domain binds preferentially to PI ( 4 , 5 ) P2 , PI ( 4 ) P and brain lipids ( Figure 3B ) [12] . By contrast , the UNC-104 PH domain with the D1497N residue greatly reduces the preferential affinity for PI ( 4 , 5 ) P2 , PI ( 4 ) P and brain lipids ( Figure 3B , 3C ) . However the binding to both PC and PI increases compared to the wild type PH domain ( Figure 3B ) . This suggests that the D1497N PH domain variant likely retains the ability to bind lipids even though the preferential binding to PI ( 4 , 5 ) P2 is highly decreased . For further confirmation , we tested whether increasing PI ( 4 , 5 ) P2 in vivo provided functional rescue . In the transgenic line gqIs25 , which over-expresses the PI ( 4 , 5 ) P2 biosynthetic enzyme ppk-1 in neurons , PI ( 4 , 5 ) P2 levels are increased by 40% in vivo [35] . We tested functional rescue of transport using a locomotory behavioural assay and an aldicarb resistance assay . These assays depend on the release of neurotransmitter filled vesicles at synapses [36] that have been transported by the UNC-104 motor . ( See materials and methods for the inverse relationship between synaptic transmission and paralysis induced by the acetylcholine esterase inhibitor aldicarb . ) In unc-104 mutants , vesicles at synaptic regions are greatly reduced [3] , resulting in animals that are nearly immobile due to reduced synaptic transmission and are greatly resistant to paralysis induced by aldicarb ( Figure 3D , 3E ) [10] . Wild type and PI ( 4 , 5 ) P2 over-expressing animals have robust locomotion and are highly sensitive to aldicarb ( Figure 3D , 3E ) . In unc-104 ( e1265 ) animals over-expressing PI ( 4 , 5 ) P2 , there is no improvement in locomotory behaviour or sensitivity to aldicarb when compared to unc-104 ( e1265 ) animals ( Figure 3D , 3E ) . Thus the protein encoded by the unc-104 ( e1265 ) allele with the D1497N lesion is insensitive to PI ( 4 , 5 ) P2 levels in vivo . We wished to confirm that the reduced ability of the D1497N PH domain to bind PI ( 4 , 5 ) P2 in vitro results in correspondingly reduced ability of the UNC-104 ( D1497N ) motor to bind to its synaptic vesicle cargo in vivo . For this we prepared pre-synaptic vesicles from unc-104 ( e1265 ) and wild type animals . Both genotypes have nearly identical levels of synaptic vesicles as assayed by the vesicle marker synaptobrevin ( Figure 2E ) . At the same time , the vesicles prepared from unc-104 ( e1265 ) animals have very low amounts of UNC-104 present on them when compared to vesicles prepared from wild type animals ( Figure 2E ) . Thus UNC-104 ( D1497N ) , encoded by unc-104 ( e1265 ) , loses its ability to bind to PI ( 4 , 5 ) P2 in vitro , has greatly reduced presence of the mutant motor on its vesicular cargo in vivo and shows loss of synaptic vesicle transport irrespective of PI ( 4 , 5 ) P2 levels in vivo . This suggests that the motor encoded by the unc-104 ( e1265 ) allele is unable to transport synaptic vesicles through its inability to bind to its cargo . To ameliorate the effects of unc-104 ( e1265 ) and to improve in vivo cargo binding , we screened ∼40 , 000 genomes in a behavioural suppressor screen using EMS mutagenesis . We identified four independent intragenic suppressors within the PH domain that improved the locomotory behaviour of unc-104 ( e1265 ) ( Figure 3D ) . Of these , three alter the same residue , M1540I , while the other suppressor was an alteration at the highly conserved residue R1501 to Q1501 ( Figure S1A , S1B ) . The three hits that result in the M1540I change were independently isolated at different times and did not always have the same nucleotide change ( Figure S1A ) . We wished to determine if the suppressors , unc-104 ( e1265tb107 ) and unc-104 ( e1265tb120 ) ( respectively encoding the proteins UNC-104 ( D1497N R1501Q ) and UNC-104 ( D1497N M1540I ) ) , improved behaviour by altering synaptic vesicle distribution . We carried out an aldicarb resistance assay and also directly observed the distribution of a synaptic vesicle protein in motor neurons . The intragenic suppressors are less resistant to aldicarb compared to unc-104 ( e1265 ) ( Figure 3E , Figure S2C ) . This indicates that both intragenic suppressors have greater release of acetylcholine at synapses . Consistent with these observations , we found that synaptobrevin-1::GFP ( SNB-1::GFP ) , a synaptic vesicle protein marker transgenically expressed in motor neurons , accumulates largely in cell bodies rather than at synapses of unc-104 ( e1265 ) animals ( Figure 4A: b ) [24] , [37] and the number of muscle arms connecting with synapses is greatly reduced ( Figure S2A , S2B ) [38] . In both intragenic suppressors the accumulation of SNB-1::GFP in cell bodies is greatly reduced and correspondingly more SNB-1::GFP is present at synapses and the number of muscle arms increases significantly ( Figure 4A: c , d , Figure S2A , S2B ) . Another synaptic vesicle marker GFP::RAB-3 [27] , [28] behaves identically to SNB-1::GFP in mechanosensory neurons . In unc-104 ( e1265 ) the marker GFP::RAB-3 accumulates in the cell body with nearly no protein present at synapses . In both intragenic suppressors more GFP::RAB-3 is present in synaptic regions with lower accumulations in the cell body ( Figure S2D ) . Consistent with this increase of GFP::RAB-3 in synaptic regions , the anterograde flux of GFP::RAB-3 in mechanosensory neurons is higher in the intragenic suppressors than in unc-104 ( e1265 ) ( Videos S2 , S3 ) , but still significantly less than in wild type ( Figure 4D , 4E , Video S1 ) . There is a reduction in the anterograde velocity of GFP::RAB-3 only in unc-104 ( e1265 ) animals ( Figure 4F ) , suggesting that in the suppressors , the partially functional UNC-104 motors that succeed in binding cargo are able to transport it efficiently . The retrograde flux of GFP::RAB-3 is also reduced in all three mutants , greatly so in unc-104 ( e1265 ) but only moderately in the suppressors ( Figure 4D ) . The retrograde velocity is unaffected in all mutants ( Figure 4F ) , suggesting that the reduced retrograde flux is likely due to fewer cargo vesicles being available for retrograde transport as a result of reduced anterograde transport . Thus both intragenic suppressors that map to the PH domain improve behaviour and cholinergic synaptic transmission by increasing the transport of cargo in the axon and the number of synaptic vesicles at the synapse . To determine if the intragenic suppressors improve synaptic vesicle transport through improved cargo binding , we carried out a parallel analysis of the suppressors in a manner similar to unc-104 ( e1265 ) . The R1501Q mutation in unc-104 ( e1265tb107 ) is also on the surface of the PH domain ( Figure 3A ) . It lies ∼7 Å away from D1497N within the PI ( 4 , 5 ) P2 binding pocket . The R1501Q may reverse the loss of charge in the PH domain variant encoded by unc-104 ( e1265 ) through compensatory local short range interactions . The compensatory change M1540I in the unc-104 ( e1265tb120 ) suppressor lies ∼20 Å from D1497N and is likely to mediate any possible effect on PI ( 4 , 5 ) P2 through long-range interactions ( Figure 3A ) . In vitro lipid binding using the D1497 R1501Q and D1497N M1540I PH domains showed that they partially restore preferential PI ( 4 , 5 ) P2 binding ( Figure 3B , 3C ) . D1497N R1501Q shows a small increase in PI ( 4 ) P binding but D1497N M1540I does not show a similar increase . These suppressor variants of the UNC-104 PH domain , like the D1497N variant , continue to bind PC and PI at higher levels than the PH domain encoded by wild type ( Figure 3B ) . These data suggest that along with partial restoration of preferential binding to PI ( 4 , 5 ) P2 , some non-specific binding to lipids is still retained by the PH domains encoded by the intragenic suppressors . As a confirmation that the intragenic suppressors are able to recognize PI ( 4 , 5 ) P2 in vivo , we observed that increased PI ( 4 , 5 ) P2 resulting from neuronal over-expression of ppk-1 [35] leads in each suppressor to improved locomotion and reduced resistance to aldicarb induced paralysis ( Figure 3D , 3E ) . This indicates increased transport to synapses resulting in increased vesicle release in the intragenic suppressors over-expressing PI ( 4 , 5 ) P2 . The previously described [12] engineered mutants KK1463/4AA and R1496A are also similarly sensitive to PI ( 4 , 5 ) P2 levels in vivo ( Figure S3E ) , suggesting that they do not reduce PI ( 4 , 5 ) P2 binding as severely as the D1497N variant . Consistent with the in vitro and in vivo data , we observe that pre-synaptic vesicles prepared from unc-104 ( e1265tb120 ) animals have larger amounts of UNC-104 present on them than those prepared from unc-104 ( e1265 ) animals ( Figure 2E ) . Both genotypes have nearly identical levels of synaptic vesicles as assessed by levels of the synaptic vesicle marker synaptobrevin ( Figure 2E ) . Thus , compared to unc-104 ( e1265 ) , the proteins encoded by the intragenic suppressors ( 1 ) partially restore preferential PI ( 4 , 5 ) P2 binding in vitro , ( 2 ) are sensitive to PI ( 4 , 5 ) P2 levels in vivo ( 3 ) have more UNC-104 molecules on pre-synaptic vesicles and ( 4 ) facilitate transport of synaptic vesicles to synapses through an improved ability to bind cargo vesicles , leading to improved behaviour . To investigate consequences of cargo binding ability on the motor we examined the levels of the pan-neurally expressed UNC-104 motor in several alleles ( Figure S4D ) . Greatest levels of endogenous UNC-104 are found in the synapse rich regions of the nerve ring and of the ventral cord ( Figure S4A:a , Figure 4B ) . Lower levels of UNC-104 are present in the dorsal cord , sub-lateral cords and in neuronal commissural processes ( Figure S4B ) . UNC-104 levels in unc-104 ( e1265 ) are greatly reduced compared to wild type animals and residual protein is still localized in the synapse rich regions of the nerve ring and ventral cord ( Figure 4B , 4C:b , g , Figure S4A: c ) . As a comparison no change was observed in the levels or localization of the neuronal plasma membrane t-snare syntaxin ( Figure 4C: k , l ) . In another pre-existing allele unc-104 ( rh43 ) , which encodes the motor UNC-104 ( G96E G314E ) with a mutation in the ATP binding pocket of the motor domain ( Figure S1A ) , the UNC-104 levels appear similar to wild type , although altered in distribution with significant increases in neuronal cell bodies ( Figure 4C: c , h , Figure S4A: b ) . The altered distribution may arise from a motor that is unable to hydrolyze ATP and thus cannot walk efficiently along microtubules . To see how partial restoration of the pattern that favours PI ( 4 , 5 ) P2 binding affects UNC-104 levels , we carried out immunohistochemistry and Western blots on both intragenic suppressors . We observed that the UNC-104 protein levels are also partially restored in the intragenic suppressors ( Figure 4B , 4C: d , e , i , j ) . Moreover , this increase occurs where the endogenous levels of UNC-104 were highest , namely in the synapse rich regions of the nerve ring and of the ventral cord ( Figure 4C: d , e , i , j ) . These regions also contain axons , so some of the increase could be taking place in axons . Upon increasing the in vivo levels of PI ( 4 , 5 ) P2 in intragenic suppressors , along with improved behaviour ( Figure 3D , 3E ) , we see a further increase in UNC-104 levels ( Figure S3F ) . Again this additional increase in UNC-104 levels is detected only in the synapse rich regions of the ventral cord ( Figure S4I ) . This is likely due to an increased number of partially functional motors being recruited to cargo vesicles . ( The above data are summarized in Table 1 ) Taken together , our observations show that the in vivo levels of the UNC-104 motor are directly related to its ability to bind pre-synaptic vesicles through PI ( 4 , 5 ) P2 , suggesting a link between specific binding of a motor to its cargo and levels of the motor in neurons . We wished to test if the reduced UNC-104 levels in the unc-104 variants are due to its degradation . To rule out reduction in transcripts , we measured RNA levels of UNC-104 using real-time PCR . We saw no change in UNC-104 RNA levels between wild type and unc-104 ( e1265 ) animals ( Figure S3D ) . To study other possible effects of the D1497N mutation on the UNC-104 motor such as altered localization or motility , we compared transgenic animals expressing high levels of UNC-104::GFP and UNC-104 ( D1497N ) ::GFP . High levels were used since at low levels , there is almost no expression of the mutant motor in vivo . Both variants show similar localization and nearly identical microscopic movements ( Figure 4G , see below ) . Nearly 85% of UNC-104::GFP and 75% of UNC-104 ( D1497N ) ::GFP molecules that move do so in the anterograde direction while approximately 15-25% move in the retrograde direction ( Figure 4G ) . Thus mis-localization or immobility of the UNC-104 motor are unlikely to underlie the observed phenotypes of unc-104 ( e1265 ) animals . To test if UNC-104 is degraded in the unc-104 allelic variants we built double mutants between these variants and the temperature sensitive allele of the E1 Ubiquitin ligase uba-1 ( it129ts ) [26] . We observed a small but consistent increase in UNC-104 levels on Western blots in unc-104 ( e1265 ) ; uba-1 animals grown at 22°C compared to unc-104 ( e1265 ) animals grown at the same temperature ( Figure 5A1 ) . This increase , observed primarily in the nerve ring ( Figure 5B: b , f , j , n ) , did not result in any improvement in resistance to aldicarb ( Figure 5E ) , probably because the mutant motors are still unable to bind cargo efficiently for transport . Similar results were obtained for the two intragenic suppressors . Western blots showed increased UNC-104 levels in each suppressor in the uba-1 background ( Figure 5A2 ) . Again this increase occurs in the synapse rich regions of the nerve ring and of the ventral cord ( Figure 5B: c , g , k , o , d , h , l , p ) . These regions also contain axons , so the increase in motors may occur to some degree in axons in addition to synapses . Concomitant with the increase in motor levels , we observed a significant increase in synaptic vesicles at neuromuscular junction synapses in unc-104 ( e1265tb107 ) and unc-104 ( e1265tb120 ) in the uba-1 background ( Figure 5C , 5D ) . This was reflected in better behaviour , namely we saw greater sensitivity to aldicarb at 22°C in both suppressors in the uba-1 background ( Figure 5E ) . These data indicate that blocking the ubiquitin-mediated degradation pathway in the suppressors increases the numbers of partially functional motors , which likely improves the transport of pre-synaptic vesicles , resulting in improved synaptic transmission . Taken together , our observations suggest that loss of ability to bind cargo can lead to motor degradation in neurons . Further , the UNC-104 motors that have reduced binding ability to PI ( 4 , 5 ) P2 are degraded at least partially through the ubiquitin pathway in synapse rich regions of the nerve ring and ventral cord . To provide further support for the observed loss of the UNC-104 motor upon lack of PI ( 4 , 5 ) P2 binding , we made low copy number transgenic lines of several UNC-104 variants by bombardment into unc-104 ( e1265 ) animals . The UNC-104 motor variant transgenic lines were made using wild type UNC-104 , UNC-104 ( D1497N ) , UNC-104 ( D1497N R1501Q ) , UNC-104 ( D1497N M1540I ) with and without a GFP fused to the C-terminus expressed under the control of the unc-104 promoter . The GFP containing and GFP lacking transgenic lines behaved identically in both locomotion and aldicarb resistance assays , suggesting that addition of GFP did not alter the function of the UNC-104 variants ( Figure S3B1 , S3B2 , S3C1 , S3C2 ) . All variants except UNC-104 ( D1497N ) provided full or partial rescue of the localization of GFP::RAB-3 in a pattern similar to that observed in wild type animals ( Figure S4E ) . All transgenic lines except UNC-104 ( D1497N ) ::GFP provide significant restoration of both locomotion and synaptic transmission as assayed by aldicarb sensitivity ( Figure 6B , 6C ) . All transgenic lines except UNC-104 ( D1497N ) ::GFP express GFP in a pattern similar to the pattern of immunoreactivity seen in wild type animals ( Figure 6A ) . However none of the three independently generated UNC-104 ( D1497N ) ::GFP transgenic lines express GFP ( Figure 6A: c1 , c2 ) . Further , injecting the UNC-104 ( D1497N ) ::GFP construct at high DNA concentrations ( ∼200ng/µl ) did result in motor-GFP expression in a pattern similar to high copy number UNC-104::GFP transgenic lines ( Figure 6A: b1 , b2 , d1 , d2; Figure S4C ) . We think that this expression in very high copy number transgenic UNC-104 ( D1497N ) ::GFP lines is likely due to saturation of the endogenous degradation machinery . These transgenic animals clearly demonstrate that some fusion protein expressing GFP could be produced by this construct when sufficiently high copy numbers of the encoding DNA are provided but this expression still does not provide behavioural rescue ( data not shown , all transgenic data are summarized in Table 1 ) . However when expressed at levels closer to endogenous levels , the UNC-104 ( D1497N ) variant is not detectable , possibly due to being targeted for degradation . The most parsimonious conclusion is that specific binding to PI ( 4 , 5 ) P2 molecules present on cargo vesicles is essential for maintaining the levels of the UNC-104 motor . To further confirm that the ability to maintain preferential PI ( 4 , 5 ) P2 binding is co-related to in vivo motor levels , we mutated the W1549 to A . The intragenic suppressor M1540I carries out its suppression indirectly . This residue is ∼2 . 4 Å away from the highly conserved Trptophan at 1549 . In the homology model , I1540 orients its β carbon methyl group towards W1549 , which in turn lies close to KK1463/4 ( ∼7 . 8 Å/3 . 2 Å respectively ) ( Figure 3A ) . We predicted that the W1549 residue would mediate the suppression of M1540I through interaction with the classical KK1463/4 residues . KK1463/4 are known to play important roles in vitro in binding PI ( 4 , 5 ) P2 [12] . Since the side chain of isoleucine is bulkier than methionine , M1540I mutation might be amenable for better interaction with the conserved W1549 . This may directly cause a change in the binding site for better presentation to the ligand . Thus , changing the W1549 to A1549 is likely to reduce the presumptive interaction from I1540 to the KK1463/4 . Therefore , we tested the in vitro lipid binding specificity of an UNC-104 PH domain carrying ( D1497N M1540I W1549A ) mutations and observed that this triple mutation abolishes the preferential PI ( 4 , 5 ) P2 binding in vitro while increasing binding to PC and PI , and behaves similarly to the D1497N mutation alone ( Figure 3B , 3C ) . As predicted , this triple mutation abrogates the ability of the M1540I to suppress the deleterious effects of the D1497N lesion . We also made low copy number integrated transgenic lines using bombardment with both UNC-104 ( W1549A ) ::GFP and UNC-104 ( D1497N M1540I W1549A ) ::GFP into unc-104 ( e1265 ) . The UNC-104 ( W1549A ) ::GFP lines exhibit wild type locomotory behaviour , sensitivity to aldicarb and motor expression levels and localization ( Figure 6A: j1 , j2 ) , suggesting that a motor with W1549A does not materially alter function in vivo ( Figure 6B , 6C , Figure S3B1 , S3B2 ) . However the UNC-104 ( D1497N M1540I W1549A ) ::GFP does not show any expression of GFP in any of the transgenic lines generated ( Figure 6A:f1 , f2 ) . Nor does it exhibit normal locomotion and moreover the unc-104 ( e1265 ) animals carrying this transgene continue to be resistant to aldicarb ( Figure 6B , 6C ) . Thus all analyzed mutations that reduced the preferential PI ( 4 , 5 ) P2 binding also show reduced UNC-104 motor protein levels .
Failure of UNC-104 to return from synaptic regions in C . elegans neurons ( Figure 2 A-C ) corroborates prior reports showing , via axon ligation assays , that motors such as the mammalian KIF1A , Kinesin-1 and KIF3A/B do not get retrogradely transported back to the cell body [6] , [21] , [22] . Degradation at synapses can explain this apparent macroscopic directionality of anterograde transport . Such degradation could have consequences for cargo transport , for instance , by providing a mechanism for preventing tug-of-war with a retrograde motor or return of retrogradely directed cargo back to the synapse . That UNC-104 is degraded near synaptic regions is demonstrated by increase in motor levels at synapses in mechanosensory neurons upon blocking ubiquitin-mediated degradation ( Figure 1A:e1 , e4 ) . Together with the observed in vivo ubiquitination of UNC-104 ( Figure 2D ) , this suggests that degradation of the motor at synapses is mediated directly or indirectly by ubiquitination . The degradation of UNC-104 through the ubiquitin pathway is likely to require the PH domain since a transgenic motor::GFP fusion protein lacking the PH domain has been shown to be highly expressed [12] ( Figure S4C ) . This explanation is also consistent with the observed direct interaction of ubiquitin with a split PH domain that shares significant homology to the UNC-104 PH domain [39] . Further , the UNC-104 PH domain has 70% similarity to a 43 amino acid ubiquitin-mediated degradation sequence found in kinesin Kip1p [40] . Moreover several lysine residues are present in the PH domain , including three in the PH domain that may be targets for attaching ubiquitin to the UNC-104 motor ( Figure S1B ) . While these facts suggest that the degradation of UNC-104 is likely to occur via ubiquitin interactions with the PH domain of the motor , we cannot rule out other degradation pathways , for instance involving a more indirect role for ubiquitination and/or a role for phosphoinositides [41] . In the allelic series consisting of wild type , unc-104 ( e1265 ) and its two intragenic suppressors , the ability to bind PI ( 4 , 5 ) P2 determines the levels of motors on pre-synaptic vesicles in vivo , the extent of transport of synaptic vesicle proteins to the synapse , and hence the extent of locomotion and of synaptic transmission . We think that the primary defect in these unc-104 mutants is differential abrogation of cargo binding ability , rather than other effects such as altered localization , motility , folding or stability of the mutant motor . The fact that over-expressed UNC-104 ( D1497N ) ::GFP and over-expressed UNC-104::GFP localize and move similarly in vivo ( Figure 6A , Figure 4G ) argues against localization and motility being affected . We cannot currently exclude the possibility that protein folding or stability is changed in vivo . We discuss this in the next section . We also found the levels of UNC-104 in all alleles to be directly related to their cargo binding ability . Further the mutant motors undergo ubiquitin-mediated degradation in synapse rich regions of the animal , as seen by the small increase ( see the next paragraph ) in UNC-104 expression in these regions after blocking ubiquitin-mediated degradation ( Figure 5B ) . The D1497N lesion in itself is unlikely to cause the mutant UNC-104 motor to be targeted for degradation since the D1497N residue is not a direct target for ubiquitin conjugation , and hence UNC-104 ( D1497N ) is unlikely to generate a new site for poly-ubiquitin attachment . The likely reason why only a small increase is observed in mutant UNC-104 levels upon blocking ubiquitination is that uba-1 is a mild temperature sensitive mutant providing sufficient function for viability of the uba-1 animals . This is also the probable reason behind the apparent lack of change seen in endogenous UNC-104 levels in uba-1 mutants alone in these assays ( Figure 5A1 , 5B: i , m ) . One would expect to see such a change in view of the independently established degradation of the UNC-104 motor in mechanosensory neurons ( Figure 1A:e1 , e4 ) . But since endogenous wild type UNC-104 is present in all neurons in large amounts , we think that the small change caused by uba-1 is difficult to detect . It may be possible to see more robust effects , including on endogenous wild type UNC-104 , if one identifies a specific E3 ubiquitin ligase , rather than using a general block of degradation provided by uba-1 . The observed degradation patterns of UNC-104 in the animals in the allelic series , coupled with the direct relationship between UNC-104 levels in these animals and cargo binding ability of the mutant motors , provide support to the following hypothesis . The endogenous UNC-104 motor carrying synaptic vesicles goes to synaptic regions and is degraded there upon cargo release . At present we cannot rule out potential instability of the mutant UNC-104 motor as the primary factor leading to its degradation and hence to loss of cargo binding and other ensuing phenotypes . However this explanation is considerably less parsimonious since it leaves unexplained the following localization and movement patterns of mutant motors . Observed steady state localization of all three mutant motors is confined to synapse rich regions , as is the increase upon blocking ubiquitination ( Figure 5B ) . This suggests that the mutant motors can get transported to synaptic regions . Further , the microscopic movements of over-expressed UNC-104 ( D1497N ) ::GFP suggests that at least some mutant motors are able to fold and move correctly ( Figure 4G ) . Moreover , the nearly identical CD melting spectra of both the wild type and D1497N PH domains imply their structural similarity ( data not shown ) . The fate of UNC-104 motors not carrying synaptic vesicles is less clear . In case of the mammalian KIF1A , motors unbound to cargo have recently been shown to be held in an auto-inhibited state preventing transport to neurite tips [17] . However there are reported differences between UNC-104 and KIF1A , e . g . , UNC-104 appears to exist as a monomer and is thought to dimerize on the surface of the cargo [14] , whereas KIF1A has been reported to move as a monomer [42] and recently it is reported to be held as a dimer [17] . Moreover UNC-104 has been previously reported to enter axons even after deletion of its cargo binding PH domain , demonstrating that cargo binding is not necessary for movement of the motor ( Figure S4C ) [12] . In all our allelic variants including wild type , we find almost no UNC-104 present in most neuronal cell bodies , even in the uba-1 background ( Figure 5B: e , m , o , p contrast with Figure 4C: h ) . One possible explanation is that most motors enter the axon very quickly – with mutant versions conceivably carrying other lipids or even no cargo – and upon reaching synaptic regions and after losing binding to cargo , the motor is rapidly targeted for degradation . Other mechanisms , such as degradation of motor as soon as the motor-cargo complex reaches the synapse or only after inactivation of the motor , cannot be excluded . However , our work suggests that a plausible mechanism is one in which release of the motor from its cargo may expose free motors to degradation at or near the synapse .
A BLAST search of the UNC-104 PH domain sequence against the RSCB protein data bank identifies DAPP1/PHISH ( Dual adaptor of phosphotyrosine and 3-phosphoinositides , from Homo sapiens , PDB code 1FB8 ) as the closest homolog . The two sequences were then aligned using CLUSTAL W ( EBI server ) and carefully adjusted using manual intervention , to ensure maximum conservation of motifs and minimal gap regions . A homology model was then generated using MODELLER v7 . 0 [43] . Output structure was relaxed with 500 steps of energy minimization ( Steepest Descent ) using SYBYL ( Tripos Associates , Inc . ) . The energy-minimized structure was then used as input for docking PI ( 4 , 5 ) P2 using GRAMM [34] . The starting constructs for all was an UNC-104 PH domain fused in frame to GFP [25] . Various point mutations ( D1497N , D1497N M1540I , D1497N R1501Q , D1497N W1549A M1540I ) in the PH domain were generated using site directed mutagenesis using the Stratagene QuickChange protocol with TaKaRa Ex Taq . PH domain constructs were cloned into pET17b vector and all constructs were verified by DNA sequencing . The proteins were expressed in Rosetta bacterial cells ( Invitrogen ) , purified by Ni-NTA chromatography ( QIAGEN ) and kept frozen in 10mM Tris pH 8 . 0 , 4mM EGTA , 5% sucrose . The followings lipids were purchased from Avanti Polar lipids . Egg PC ( Cat . no . 840051 ) , PI ( 4 ) P ( Cat . no . 840045 ) , PI ( 4 , 5 ) P2 ( Cat . no . 840046 ) , PA ( Cat . no . 840101 ) , PI ( Cat . no . 840044 ) and BL ( Cat . no . 131101 ) . Composition of brain lipids ( BL ) contains Phosphatidylethanolamine ( 16 . 7% ) , Phosphatidylserine ( 10 . 6% ) , Phosphatidylcholine ( 9 . 6% ) , Phosphatidic acid ( 2 . 8% ) , Phosphatidylinositol ( 1 . 6% ) and others ( 58 . 7% ) . 5 µM concentration of the desired lipids was used to prepare liposomes in the following ratio: 10% desired lipid and 90% carrier lipids . Phosphatidylcholine ( PC ) was used as a carrier lipid and the remaining 10% of the lipids used were either Phosphatidylinositol ( PI ) , Phosphatidylinositol-4-phosphate PI ( 4 ) P , Phosphatidylinositol-4 , 5-bisphosphate PI ( 4 , 5 ) P2 or brain lipid ( BL ) in chloroform . After mixing the desired and carrier lipids , chloroform was evaporated under a constant Nitrogen gas stream . Once the lipid film was dried completely , lipids were rehydrated by the addition of LB buffer ( 30 mM tris , 4 mM EGTA , pH 8 . 0 ) . These lipids were sonicated ( ultrasonic bath ) for 30 seconds to break up the lipid aggregates and were then extruded through a 100 nm pore polycarbonate filter ( Avestin , Ottawa , Canada ) using a miniextruder from Avanti polar lipids . The liposomes were stored in the dark at 4°C and used within a week of preparation . Liposomes were prepared as previously described [14] . Briefly , liposomes ( 5 µM total lipid concentration ) were prepared in LB buffer ( 30 mM tris , 4 mM EGTA , pH 8 . 0 ) . 100 µl of freshly prepared liposomes were mixed with about 1 µg protein and incubated on ice for 30 min . The incubation reaction mixtures were centrifuged at 50000gav ( 4°C ) for 45 min in a TLS-120 rotor ( Beckman ) . After centrifugation , fractions from the pellet that contains liposome bound protein and supernatant that contains unbound protein were collected . Samples were dissolved in 20 µl LB buffer and analyzed by SDS-PAGE followed by Coomassie staining . Gels were digitized on a flatbed scanner and protein bands were quantified using ImageJ ( version 1 . 37 , NIH ) . Binding specificity was determined by normalizing binding observed with PI ( 4 , 5 ) P2 and brain lipid compared to binding observed using PC alone carrier liposomes . L4 unc-104 ( e1265 ) worms were washed with M9 buffer , using sterile glass pipettes . Washed worms were transferred into a tube of 1x PBS containing ethyl methanesulfonate ( Sigma ) at a final concentration of 50mM . Tubes were kept in a rotary shaker at 20°C for 4 hours . After mutagenesis , 3-4 worms were transferred each 60 mm Petri plate . F1 and F2 progeny were regularly examined under a Nikon SMZ645 dissecting microscope for improved locomotion in a non-clonal screen of approximately 60 , 000 haploid genomes . Intragenic suppressors were identified in genetic crosses that mapped them close to the unc-104 locus . Intragenic suppressors of unc-104 ( e1265 ) isolated were unc-104 ( e1265tb107 ) and unc-104 ( e1265tb120 ) . Throughout the paper , proteins encoded by these alleles are referred to as UNC-104 ( D1497N ) , UNC-104 ( D1497N R1501Q ) and UNC-104 ( D1497N M1540I ) respectively . 1 day adult hermaphrodites were transferred on to a fresh NGM agar plate , allowing them to acclimatize for 1 hour . Movement was recorded on a Nikon SMZ800 dissecting microscope at 1 to 1 . 3 frames per sec ( 1000×1000 pixels ) for 2-3 min with a cooled monochrome camera ( Evolution Qei , Media Cybernetics ) . Movement was tracked manually using ImageJ ( version 1 . 37 , NIH ) software . Worms that moved for a minimum of 10 frames were tracked . Worm velocities were obtained by calculating the straight line distance between the centroid positions of the worm in a given interval . Aldicarb plates were prepared by adding aldicarb ( Chemical Service , Westchester , PA ) solution ( in 70% ethanol ) to NGM agar . These plates were seeded with OP50 bacteria . All assays were performed on 1 day old adult hermaphrodites at room temperature ( 21-23°C ) . 30 individuals were incubated for 6-8 hr on aldicarb plates of defined concentration . At 30 min intervals each worm was touched with a platinum wire and was checked for paralysis [36] . Aldicarb inhibits acetylcholine esterase causing the neurotransmitter acetylcholine to persist longer at the synapse and hyperstimulate the post-synaptic sites . This leads to loss of co-ordinated motion and finally paralysis . Faster paralysis indicates more acetylcholine release at synapses . In our experimental context wild type paralyzes the fastest while mutants that do not have vesicles to release paralyze the slowest . Any reduction in paralysis time indicates more vesicles present at synapses for release . A wild type UNC-104::GFP construct was provided by Jon Scholey [25] . This construct harbours the unc-104 promoter driving the combination of intronless and genomic region of unc-104 and provides the entire open reading frame of the protein . Mutations were introduced using site directed mutagenesis using the Stratagene QuickChange protocol with TaKaRa Ex Taq . Various point mutations ( D1497N; D1497N M1540I , M1540I , D1497N R1501Q , R1501Q , W1549A , D1497N M1540 W1549A ) were generated . All constructs were verified by DNA sequencing . GFP was deleted from UNC-104::GFP , UNC-104::GFP ( D1497N ) , UNC-104::GFP ( D1497N R1501Q ) and UNC-104::GFP ( D1497N M1540I ) using the restriction enzymes Apa1 and Kpn1 . After T4 DNA polymerase treatment , ligation was done using T4 DNA ligase . Worms were grown at 20°C on NGM agar plates seeded with E . coli Strain OP50 under standard laboratory conditions ( Brenner , 1974 ) . Strains used in the study , provided by the Caenorhabditis Genetics Center ( CGC ) , are as follows: wild type N2 , unc-104 ( e1265 ) , unc-104 ( rh43 ) , unc-104 ( rh142 ) . juIs1 ( punc-25-SNB-1::GFP ) a transgenic strain expressing green fluorescent protein ( GFP ) -tagged synaptobrevin-1 in GABA motor neurons [24] , [37] jsIs1 ( psnb-1::snb-1::GFP ) a transgenic strain that expresses SNB-1::GFP in all neurons [44] . zdIs5 ( pmec4::GFP ) a transgeneic strain expressing soluble GFP in mechanosensory neurons [29] jsIs821 ( pmec7::GFP::RAB-3 ) a transgenic strain expressing GFP tagged RAB-3 in mechanosensory neurons [27] trIs25 ( him-4p::MB::YFP , F25B3 . 3P::DsRed2 ) has Membrane–anchored yellow-fluorescent protein expressed in body wall muscles [38] . jsIs682 ( prab-3::gfp::rab-3 ) a transgenic strain that expresses GFP::RAB-3 pan-neurally [27] . gqIs125 ( prab-3::ppk-1 ) a transgenic strain that over-expresses the PI ( 4 , 5 ) P2 biosynthetic enzyme Type I PIP kinase ppk-1 in all neurons [35] uba-1 ( it129ts ) is a temperature sensitive mutant allele in the E1 ubiquitin activating enzyme [26] js1111 ( pmec4::UNC-104::GFP ) a transgenic strain that expresses UNC-104::GFP only in mechanosensory neurons . UNC-104- 5 transgenic lines- tbIs183 UNC-104::GFP- 5 transgenic lines- tbIs147 UNC-104 ( D1497N ) - 3 transgenic lines- tbIs188 . This transgene rescues unc-104 ( rh142 ) , the lethal null allele UNC-104 ( D1497N ) ::GFP- 3 transgenic lines- tbIs149 . This transgene provides viability to unc-104 ( rh142 ) , the lethal null allele UNC-104 ( D1497N R1501Q ) - 3 transgenic lines- tbIs194 UNC-104 ( D1497N R1501Q ) ::GFP- 3 transgenic lines- tbIs152 UNC-104 ( D1497N M1501I ) - 4 transgenic lines- tbIs191 UNC-104 ( D1497N M1501I ) ::GFP- 3 transgenic lines- tbIs156 UNC-104 ( M1540I ) ::GFP-13 transgenic lines- tbIs157 UNC-104 ( R1501Q ) ::GFP- 4 transgenic lines- tbIs170 UNC-104 ( D1497N M1540I W1549A ) ::GFP- 5 transgenic lines- tbIs199 UNC-104 ( W1549A ) ::GFP- 2 transgenic lines – tbIs181 Underlined strain was most commonly used , at least one other transgenic strain was assayed in all assays and no co-injection marker was used to make the above transgenic animals . Micro particle bombardment of C . elegans unc-104 ( e1265 ) hermaphrodites was carried out using a BioRad Biolistic PDS-1000/HE particle delivery system ( Bio-Rad Laboratories , Hercules , CA , USA ) [45] . For each bombardment , 5-6 µg plasmid DNA was fixed to 0 . 5mg of 1 . 0 µm micro carrier tungsten particles , as described in the PDS-1000/HE user's manual , and bombarded on to a monolayer of unc-104 ( e1265 ) L4 . Worms were allowed to recover for 0 . 5 to 1 hr after bombardment and were then transferred on to 100mm seeded Na22 plates and grown at 20°C . After 8-12 days worms were screened for improved movement and/or GFP expression as examined using a Zeiss fluorescence microscope . Individual animals were cloned . Homozygous stable lines were identified by the complete absence of unc-104 ( e1265 ) mutant progeny over several generations [45] . We used unc-104 ( e1265 ) as the background for bombardment since this was the healthiest hypomorphic allele of unc-104 available . For quantitation of SNB-1::GFP puncta at motor neuron synapses synaptic , unsaturated images of immobilized worms were taken in the linear range of exposure and quantified using ImageJ ( NIH ) similar to what has been described in [46] . For in vivo live imaging , young adult hermaphrodites were immobilized with 3-5mM levamisole ( Sigma-Aldrich ) in M9 and mounted on a 2% agarose pad . Time-lapse images of anterior mechanosensory neurons expressing GFP::RAB-3 were obtained with OLYMPUS IX81 using 100X/1 . 4 NA plain Apochromat objective attached with spinning disk confocal head ( YOKOGAWA CSU22 ) equipped with EMCCD camera ( ANDORiXon-897EMCCD ) . Time-lapse images ( 512×512 pixels ) were taken at a constant frame rate of 6-7 frames per second . Image analysis was done using Image J ( version 1 . 37 , NIH ) . Kymographs were obtained from lines that were drawn along the axon from cell body towards synapse . Flux analysis was carried out within a range of 15-20 µm along the axon length , at a distance of 15-25 µm away from the cell body . Flux was calculated as number of anterogradely moving particles in a movie . Any particle static for 3 frames or with velocity less than 0 . 3 µm/s was considered as stationary . Pause frequency was calculated as the number of pauses taken by a particle for unit distance traveled ( number of pauses/total distance traveled ) . All significance was calculated using pair-wise comparisons using the Student's T-test with unequal variance . p values less than 0 . 05 were considered as significant . The protein region of UNC-104 ( amino acid 740-1117 ) was cloned into pRSETA vector ( Invitrogen ) using standard techniques . Protein was expressed in BL21 cells ( Invitrogen ) , and purified using Ni-NTA chromatography ( QIAGEN ) . Purified protein was given to Bioklone , Chennai , India to generate monoclonal antibodies . Specificity of the antibodies was tested by immunostaining unc-104 ( rh142 ) , a null allele . All monoclonal antibodies tested showed pan-neural staining in wild type animals and no staining in the unc-104 ( rh142 ) animals ( Figure S4D ) . Animals were fixed with 2% paraformaldehyde for 10 minutes at 4oC and freeze-thawed using liquid nitrogen and fixed for an additional 10 minutes at 4° C . Following this 4-5 washes with 0 . 5% BT buffer ( 20mM H3BO3 , 0 . 5% TritonX-100 , pH 9 . 5 ) and then 5-6 washes ( 1 hour each ) with 0 . 5%BTB ( BT with 2% mercaptoethanol ) were carried out . Blocking was done with PBST ( phosphate buffered saline , 0 . 5%BSA , 0 . 5% TritonX-100 , 10mM sodium azide ) . Samples were incubated two overnights with monoclonal anti-UNC-104 antibody ( 1:5 ) , washed for 4-5 times with PBST ( each of 15 minutes ) before mounting . Rabbit anti-syntaxin was used at 1:10 , 000 [47] . Appropriate secondary antibody ( 1:200 ) incubations ( anti-mouse Alexa 488 , Alexa 568 ) were done for two overnights at 4° C . Images were captured using Zeiss Axiovert inverted microscope . Images were processed with Adobe Photoshop Version 9 . 0 . Western sample of worms were prepared by sonication . After sonication , worm lysates were boiled with SDS lysis buffer and proteins were separated on SDS PAGE ( 8% acrylamide ) . Proteins were transferred to a nitrocellulose membrane ( Amersham ) , probed with a mouse serum or a mouse monoclonal antibody of anti-UNC-104 ( 1:60 ) , rabbit anti-tubulin ( 1:1 , 000 ) ( Thermo-scientific ) , rabbit anti-synaptobrevin ( 1:5000 ) [44] and rabbit anti-ubiquitin ( 1:500 ) ( Sigma-Aldrich ) followed by HRP based chemiluminescence detection ( Pierce ) . Exposure time was varied from 30 seconds to 5 minutes , scanned and intensities quantitated using ImageJ . These intensities were pooled from multiple experiments and graphed and the exposure time chosen was determined to be in the linear range for all genotypes . Worms of respective genotypes were anesthesized in 5mM levamisole . Photobleaching experiments were done on confocal Zeiss LSM-5 Live ( line scanner ) equipped with a 63X objective ( oil immersion , 1 . 4 NA ) with a 488 nm solid state laser . Images were acquired on a CCD camera at the frame rate of 4 Hz . 35-40 µm of the axon was bleached across the synaptic branch . Fluorescence recovery was quantified from the distance covered by the UNC-104::GFP signal in bleached axons at fixed times after bleaching . The fluorescent recovery along the anterograde and retrograde directions was represented as velocity in both anterograde ( recovery from cell body ) and retrograde ( recovery from synapse ) directions . All the analysis was done using ImageJ version1 . 41 ( NIH ) . N2 worms were used for immunoprecipitation . For sedimentation assays we used jsIs1 and various unc-104 mutants in the jsIs1 background . The worms and various mutants were grown on 10-15 large plates until food was exhausted . Worms were mechanically homogenized in homogenization buffer ( 15mM HEPES-NaOH pH 7 . 4 , 10 mM KCl , 1 . 5 mM MgCl2 , 0 . 1 mM EDTA , 0 . 5 mM EGTA 0 . 05 M sucrose and protease inhibitors ( Roche ) and mildly sonicated at 4°C . The final supernatant was centrifuged at 50 , 000g for 40 min in a TLA 100 . 3 rotor to clear debris and heavy membrane fractions . The supernatant was collected again and centrifuged at 175 , 000g in TLA100 . 3 rotor for 150 min . The final pellet was resuspended in homogenization buffer or IP buffer ( 20 mM HEPES , 40 mM KCl , 5 mM EGTA , 0 . 1m M EDTA , 5 mM MgCl2 with protease inhibitors ) as needed . For immunoprecipitation the high speed re-suspended pellet was incubated with specific antibody for 5-6 hrs at 4°C . Final concentration of UNC-104 antibody used was 1:10 and ubiquitin antibody ( Sigma-Aldrich ) used was 1:10 . Protein A agarose beads were added to the antigen-antibody mixture and incubated for 3-4 hours at 4°C . The beads were centrigufed , washed with IP buffer then analyzed by western blotting . A Western analysis was carried out on immunoprecipitated material using the anti-UNC-104 antibody and anti-ubiquitin antibody . The blot was first probed for UNC-104 and then stripped ( no signal was observed after stripping ) and re-probed for ubiquitin ( 1:500 ) ( Sigma-Aldrich ) . The anti-ubiquitin antibody recognized the same band detected by anti-UNC-104 . A Western analysis was carried out on immunoprecipitate obtained using the anti-ubiquitin antibody . This blot was probed using anti-UNC-104 and a band that migrates at the same size as endogenous UNC-104 was observed . For sucrose gradient density , the resuspended high speed pellets were loaded on a discontinuous sucrose gradient centrifugation ( 0 . 05 M , 0 . 6 M , 1 M and 1 . 5 M ) and centrifuged in a SW41 rotor at 60 , 000g for 120 min . Fractions were collected from top of the gradient up to the first layer ( between 0 . 05M-0 . 6M ) . The last two fractions collected were below the formed layer where no synaptic vesicle proteins were detected . Western blot analysis with exposure maintained in the linear range was carried out on the fractions collected . | The cell body and the synapse in a neuron are often separated by significant distance , which is spanned by the axon connecting the two . Transport of various cargoes along the axonal highway is very important for neuronal function . The regulation of this complex process is not well understood . Using the Caenorhabditis elegans model system , we have demonstrated for the first time the fate of a motor after it carries its cargo to the synapse from the cell body . We show that the UNC-104 motor , which carries pre-synaptic vesicles to the synapse , is degraded once it gets there . Moreover , our genetic studies show evidence that loss of cargo binding targets the motor for degradation , suggesting an attractive mechanism for the regulation of motors at the synapse . Our study opens up several further questions , such as the mechanism of motor degradation , and has significant implications for regulation of cargo transport . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"genetics",
"and",
"genomics/gene",
"function",
"neuroscience/neuronal",
"and",
"glial",
"cell",
"biology",
"cell",
"biology/cytoskeleton"
]
| 2010 | The Caenorhabditis elegans Kinesin-3 Motor UNC-104/KIF1A Is Degraded upon Loss of Specific Binding to Cargo |
Enterovirus 71 ( EV71 ) is an RNA virus that causes hand-foot-mouth disease ( HFMD ) , and even fatal encephalitis in children . Although EV71 pathogenesis remains largely obscure , host immune responses may play important roles in the development of diseases . Recognition of pathogens mediated by Toll-like receptors ( TLRs ) induces host immune and inflammatory responses . Intracellular TLRs must traffic from the endoplasmic reticulum ( ER ) to the endolysosomal network from where they initiate complete signaling , leading to inflammatory response . This study reveals a novel mechanism underlying the regulation of TLR7 signaling during EV71 infection . Initially , we show that multiple cytokines are differentially expressed during viral infection and demonstrate that EV71 infection induces the production of proinflammatory cytokines through regulating TLR7-mediated p38 MAPK , and NF-κB signaling pathways . Further studies reveal that the expression of the endosome-associated protein hepatocyte growth factor-regulated tyrosine kinase substrate ( HRS ) is upregulated and highly correlated with the expression of TLR7 in EV71 infected patients , mice , and cultured cells . Virus-induced HRS subsequently enhances TLR7 complex formation in early- and late-endosome by interacting with TLR7 and TAB1 . Moreover , HRS is involved in the regulation of the TLR7/NF-κB/p38 MAPK and the TLR7/NF-κB/IRF3 signaling pathways to induce proinflammatory cytokines and interferons , respectively , resulting in the orchestration of inflammatory and immune responses to the EV71 infection . Therefore , this study demonstrates that HRS acts as a key component of TLR7 signaling to orchestrate immune and inflammatory responses during EV71 infection , and provides new insights into the mechanisms underlying the regulation of host inflammation and innate immunity during EV71 infection .
Upon infection , viral RNAs are recognized as pathogen-associated molecular patterns ( PAMPs ) by Toll-like receptors ( TLRs ) to trigger signaling events leading to the induction of interferons ( IFNs ) and proinflammatory cytokines [1 , 2] . Most TLRs are intracellularly localized and must traffic from the endoplasmic reticulum ( ER ) to the endolysosomal network before they can respond to ligands [3–5] . Most RNA viruses ( eg . hepatitis C virus , and vesicular stomatitis virus ) activate TLR7 , which is initiated by binding of TLR7 to the myeloid differentiation factor 88 ( MyD88 ) adapter protein and interleukin-1 receptor-associated kinases ( IRAK ) and by recruiting tumor necrosis factor receptor-associated factor 6 ( TRAF6 ) , transforming growth factor 1-activated kinase-1 ( TAK1 ) and TAK1-binding protein 1/2 ( TAB1/2 ) [6 , 7] . These events subsequently activate multiple signaling cascades , mitogen-activated protein kinase ( MAPK ) , nuclear transcription factor-κB ( NF-κB ) and IFN regulatory factor 3/7 ( IRF3/7 ) , to induce the production of proinflammatory cytokines and IFNs , resulting in antiviral response and innate immunity [3 , 8] . Enterovirus 71 ( EV71 ) is a highly infectious positive-stranded RNA virus that causes hand-foot-mouth disease ( HFMD ) , meningoencephalitis , neonatal sepsis , and even fatal encephalitis in children [9] . Although EV71 pathogenesis remains largely obscure , host immune responses play important roles in the disease severity [10] . EV71 infection induces the production of many proinflammatory cytokines that play important roles in disease development [11–13] . Serum concentrations of interleukin-1 ( IL-1β ) , IL-1 receptor antagonist ( IL-1Ra ) , and granulocyte colony-stimulating factor ( G-CSF ) are upregulated in EV71-infected patients with cardiorespiratory compromise [11] . The overdose of proinflammatory cytokines produced during EV71 infection is mediated by the activation of different TLRs [12] . EV71 infection induces TLR7 and TLR8 in epithelial cells to enhance the induction of IFN-beta [14] . EV71 infection also upregulates intestinal tract TLR3 , TLR4 , TLR7 , and TLR8 mRNA expression of children with severe HFMD [15] . However , the molecular mechanism by which EV71 infection induces TLRs-mediated inflammatory responses is still largely unknown . Hepatocyte growth factor-regulated tyrosine kinase substrate ( HRS ) is a key component of the Endosomal Sorting Complexes Required for Transport ( ESCRT-0 ) complex and required for endosomal sorting of membrane proteins into multivesicular bodies , lysosomes and vacuoles [16 , 17] . HRS comprises an FYVE finger domain that facilitates HRS anchoring to the membrane and initiates its trafficking processes on endosomes [18] . Remarkably , HRS-mediated endosomal sorting benefits viral proteins trafficking in the host cell cytosol during viral envelopment and capsid secretion [19 , 20] . During TLRs trafficking from the ER to endolysosomes , the intracellular co-factor UNC93B1 is indispensable for the control of this process [21 , 22] . Similarly , HRS is required for ubiquitin-dependent TLR9 targeting to the endolysosome [4] . However , the detailed mechanism by which HRS regulates host inflammation and innate immunity during viral infection is still largely unknown . In this study , we reveal that a set of cytokines is expressed differentially during EV71 infection . Clinical investigations , animal studies , and cell culture experiments demonstrate that EV71 activates TLR7 to induce inflammatory factors via p38 MAPK and NF-κB transcriptional factors . By genetic screening and clinical analysis , we identified HRS acting as an essential regulator of the TLR7 signaling . HRS facilitates the TLR7 complex assembly by binding to TLR7 and TAB1 during EV71 infection . In this process , HRS enhanced proinflammatory cytokines and interferons production dependent on the activation of TLR7/NF-κB/p38 and TLR7/NF-κB/IRF3 signaling pathways . Thus , we identify a new key regulator of TLR7 signaling and provide new insights into the regulation of host immunity during EV71 infection .
To investigate the mechanism underlying inflammation and immunity induced by EV71 infection , we initially determined the effect of EV71 on the production of proinflammatory cytokines . Human Cytokine ELISA Plate Arrays showed that EV71 infection induced production of 12 cytokines , including 3 proinflammatory factors as main biomarkers in clinical prognostics [11] , colony stimulating factor 3 ( CSF3 ) , interleukin-1 beta ( IL-1β ) , and IL-6; repressed production of 4 cytokines; and had no effect on 14 cytokines in human monocytic THP-1 cells ( THP-1 ) ( Fig 1A and S1 Table ) . The role of EV71 in activation of CSF3 , IL-1β , and IL-6 was then confirmed . First , the production of CSF3 , IL-1β , and IL-6 was analyzed in peripheral blood of 40 EV71-infected patients and 36 healthy individuals ( Table 1 and S1A Fig ) . ELISA analyses indicated that CSF3 , IL-1β , and IL-6 were present in higher amount in EV71 positive patients compared with healthy individuals ( Fig 1B ) . Second , the expression of CSF3 , IL-1β , and IL-6 was determined in EV71-infected mice . The viral VP1 RNA and protein were detected in peripheral blood mononuclear cells ( PBMCs ) and the spleen of infected mice ( S1B and S1C Fig ) . Similar to patient serums , CSF3 , IL-1β , and IL-6 protein were significantly induced in peripheral blood of EV71-infected mice ( Fig 1C ) . Third , the production of CSF3 , IL-1β , and IL-6 was evaluated in EV71-infected human acute monocytic leukemia cells ( THP-1 ) , differentiated macrophages derived from THP-1 cells ( macrophages ) [23] ( S1D Fig ) , and human PBMCs cells . ELISA analyses revealed that CSF3 , IL-1β , and IL-6 proteins were induced by EV71 or R848 ( a TLR7 agonist ) as a stimulus control [24] , but not by mock-infection or UV-inactivated EV71 , in THP-1 , macrophages , and human PBMCs ( Fig 1D–1F ) . Taken together , we demonstrate that EV71 induces CSF3 , IL-1β and IL-6 both in vivo and in vitro . The mechanism by which EV71 activates proinflammatory cytokines was evaluated . Initially , signaling pathways required for EV71-mediated induction of CSF3 , IL-1β , and IL-6 production were evaluated . THP-1 cells were infected with EV71 and treated with signaling component inhibitors that had no obvious effect on cell viability or EV71 replication except PD98059 that caused a reduction in viral protein ( S2A and S2B Fig ) . CSF3 , IL-1β , and IL-6 mRNA levels were enhanced by EV71 , however , in the presence of SB203580 ( p38 MAPK inhibitor ) or Bay11-7032 ( NF-κB inhibitor ) , the induction of these mRNA levels were destroyed ( Fig 2A ) . In the mouse macrophage-like Raw264 . 7 cell line , the same phenomenon was observed ( S2C and S2D Fig ) . These results indicate that p38 MAPK and NF-κB are required for EV71-mediated induction of CSF3 , IL-1β , and IL-6 . Subsequently , the effect of EV71 on the expression of TLRs was determined . The expression of TLR7 and TLR3 mRNAs , but not TLR4 , TLR8 or TLR9 mRNAs , was upregulated by EV71 in THP-1 cells ( Fig 2B ) , suggesting that TLR3 and TLR7 may be involved in EV71-mediated induction of CSF3 , IL-1β , and IL-6 . Since it was previously reported that TLR3 signaling in macrophages is required for protection against EV71 infection in mice [25] , we further evaluated the role of TLR7 in EV71-mediated activation of proinflammatory cytokines . We applied shRNA ( short hairpin RNA ) specific to TLR7 ( shTLR7 ) that could reduce TLR7 expression ( S2E Fig ) . The mRNA levels of CSF3 , IL-1β , IL-6 , IL-8 , IL-12 , and CXCL-10 , but not TNFα , were enhanced by EV71 infection , in case of shTLR7 , mRNA levels of CSF3 , IL-1β , IL-6 , IL-8 , IL-12 , and CXCL-10 were significantly reduced , suggesting that TLR7 is involved in these cytokines production ( Fig 2C ) . To further explore the role of TLR7 in EV71-mediated induction of proinflammatory cytokines , mouse bone marrow-derived macrophages ( BMDMs ) were isolated from TLR7 WT mice and TLR7-/- mice [26] . As expected , TLR7 protein was only detected in BMDMs of TLR7 WT mice , but not in BMDMs of TLR7-/- mice ( Fig 2D ) . CSF3 , IL-1β , and IL-6 mRNAs were induced by EV71 in BMDMs of TLR7 WT mice , but not induced by EV71 in BMDMs of TLR7-/- mice ( Fig 2D ) . Similarly , CSF3 , IL-1β , and IL-6 proteins were also not induced by EV71 in BMDMs of TLR7-/- mice ( S2F Fig ) . These results confirmed that TLR7 is required for EV71-mediated induction of proinflammatory cytokines . Since TLR7 and NF-κB are required for the induction of CSF3 , IL-1β , and IL-6 production , we evaluated the correlation between TLR7 and NF-κB . First , NF-κB activity was induced by EV71 infection , but this induction was reduced in the presence of shTLR7 ( Fig 2E ) , as well as EV71-induced phosphorylation of endogenous IκBα ( Fig 2F ) . These results suggest that TLR7 is required for EV71-induced activation of NF-κB . Correspondingly , phosphorylation of NF-κB p65 and IκBα was enhanced in cells containing TLR7 upon EV71-infection , but not in mock-infected cells ( Fig 2G ) , indicating that TLR7 activates NF-κB during viral infection . Third , plasmids expressing TLR7 and the inactive TLR7 mutant Y892A were constructed and the expression of TLR7 and TLR7 ( Y892A ) proteins was confirmed ( S2G Fig ) . NF-κB activity was stimulated by TLR7 , but not by TLR7 ( Y892A ) , in the cells treated with R848 or infected with EV71 or SeV ( Fig 2H ) , implicating that TLR7 plays an essential role in EV71-mediated activation of NF-κB . Finally , a HEK293T cell line was generated that steadily expressed TLR7 and NF-κB . In this stable cell line , NF-κB activity was enhanced by EV71 infection and R848 stimulation , but attenuated in the presence of shTLR7 , shMyD88 , shIRAK1 , shNF-κB p65 , and shNF-κB p50 ( Fig 2I ) , confirming that TLR7 , MyD88 , and IRAK1 are involved in the activation of NF-κB during viral infection . Thus , these results reveal that EV71 activates TLR7 signaling to stimulate NF-κB , leading to induction of proinflammatory cytokines . To determine the mechanism by which EV71 activates TLR7/NF-κB , novel cellular factors involved in such activation were identified . A bioinformatics approach was employed to integrate a protein-protein interaction network in STRING databases [27] , which comprised known and predicted proteins associated with the TLR7 signaling pathway . By using this approach , we predicted that 14 out of 19 cellular factors were potential TLR7-associated proteins ( S3A Fig ) . We further conducted an RNAi mini-library screening to assess the functional integration of these factors in TLR7 signaling ( S2 Table ) . HEK293T/TLR7/NF-κB cells were transfected with an array of siRNAs targeting to these genes . NF-κB activity was stimulated by R848 , and significantly repressed by siRNAs to HRS ( siR-HRS ) ( S3B Fig ) , suggesting that HRS is involved in regulation of TLR7/NF-κB . Requirement of HRS for TLR7 signaling was further evaluated by using siRNA , which attenuated HRS expression ( S3C Fig ) , but had no effect on cell viability or cell apoptosis ( S3D and S3E Fig ) . NF-κB activity was upregulated in HEK293T/TLR7/NF-κB cells stimulated with R848 or TNFα or infected with EV71 or SeV , whereas the activation mediated by R848 stimulation , EV71 or SeV infection was reduced in the presence of siR-HRS , but TNFα-mediated activation was not affected in the presence of siR-HRS ( Fig 3A ) . Since EV71 activates TLR7 signaling and HRS participates in this signaling , the effect of EV71 on HRS production was investigated . TLR7 and HRS mRNAs were significantly induced in EV71-infected patients and mice PBMCs ( Fig 3B and 3C ) and TLR7 and HRS mRNAs and proteins were stimulated in EV71-infected THP-1 cells ( Fig 3D ) , demonstrating that HRS production is induced by EV71 infection in patients , mice , and cultured cells . Since EV71 induces TLR7 and HRS , the correlation between TLR7 and HRS during viral infection was determined . TLR7 and HRS mRNA levels were correlated well in EV71-infected patients and mice PBMCs ( Fig 3E and 3F ) , revealing that TLR7 production is positively correlated with HRS expression during EV71 infection . The effect of EV71 infection on HRS expression was explored . By using bioinformatic approaches , we predicted several NF-κB subunit binding sites in both human and mouse HRS and TLR7 promoters ( S4A Fig ) . In Raw264 . 7 cells , EV71 enhanced both HRS and TLR7 protein expression , but this induction was abrogated in the presence of NF-κB inhibitor Bay11-7032 ( S4B Fig ) . Considering that EV71 infection facilitates TLR7-mediated NF-κB activation , we determined whether TLR7 plays a role in EV71-induced HRS expression . EV71 facilitated HRS expression in BMDMs of TLR7 WT mice , but failed to enhance HRS expression in BMDMs of TLR7-/- mice ( S4C Fig ) , indicating that EV71 induces HRS production through TLR7 signaling . Taken together , our results suggest that HRS expression is induced by EV71 and the levels of TLR7 and HRS are correlated in vitro and in vivo upon EV71 infection . HRS acts as a mediator in endosome trafficking by binding to targeted proteins [28] and is required for TLR7-and TLR9-dependent innate immune responses [4] . To identify novel HRS-interacting factors involved in TLR7 signaling , HEK293T cells were co-transfected with plasmids expressing HRS-Myc and individual GFP-tagged proteins . Co-immunoprecipitation ( Co-IP ) showed that HRS interacted with TLR7 , TAB1 , and TAK1 , but not MyD88 , IRAK1 , TRAF3 , TRAF6 , or TBK1 ( Fig 4A ) . Interactions of HRS with TLR7 , TAB1 , and TAK1 were verified by GST pull-down assays with purified GST and GST-HRS proteins ( Fig 4B and 4C ) . To confirm such interactions of HRS with TLR7 and TAB1 , HEK293T cells were co-transfected with pHRS-Myc and pFlag-TLR7 or pFlag-TAB1 . Co-IP results further confirmed that HRS interacted with TLR7 and TAB1 ( Fig 4D and 4E ) . The associations of HRS with TLR7 and TAB1 were also examined by confocal microscopy . TLR7 and TAB1 were distributed diffusely in the cytoplasm in the absence of HRS , but co-localized with HRS and redistributed to loci in the cytoplasm in the presence of HRS ( Fig 5A ) . We also showed that the interaction of HRS with TLR7 and with TAB1 was occurred in the endosomes ( as indicated by EEA1 , an endosome marker ) , but not in the endoplasmic reticulum ( as indicated by Calnexin , an ER marker ) [29] ( Fig 5B and 5C ) . The interactions of HRS with TLR7 and TAB1 were further confirmed by bimolecular fluorescence complementation ( BiFC ) analysis in living cells ( Fig 5D ) . Fluorescence was not detected in control experiments when only one fusion protein was expressed with the complementary YFP fragment alone . Strong fluorescence was observed between VC-155-HRS with VN-173-TLR7 or VC-155-HRS with VN-173-TAB1 fusion proteins ( Fig 5E ) , suggesting that direct interactions occurred between HRS and TLR7 or between HRS and TAB1 . These results demonstrate that HRS interacts and co-localizes with TLR7 and TAB1 . To determine the function of HRS in TLR7 signaling , we evaluated the associations between HRS and the components of the TLR7 complex in differentiated macrophages derived from THP-1 cells . We found that HRS , together with MyD88 , IRAK1 , TAB1 , TRAF6 , and PKCζ proteins , could be coimmunoprecipitated with TLR7 in R848-stimulated ( Fig 6A ) or EV71-infected ( Fig 6B ) macrophages , suggesting that HRS may be a component of the TLR7 complex . To confirm the involvement of HRS in TLR7 signaling , distributions of HRS and TLR7 in macrophages were visualized by confocal microscopy . Endogenous TLR7 and HRS ( Fig 6C ) as well as endogenous TAB1 and HRS ( Fig 6D ) were distributed diffusely in the cytoplasm , but large proportions of TLR7 and HRS co-localized at foci in specific areas in the cytoplasm of R848-stimulated , EV71-infected or SeV-infected macrophages . Additionally , TLR7 mainly co-localized with EEA1 ( an early endosome marker ) and Rab7 ( a late endosome marker ) [29] in R848-stimulated macrophages ( Fig 6E ) . These results reveal that HRS co-localized with TLR7 and TAB1 during viral infection , and it most likely locates to the endosomes . Moreover , we evaluated the colocalization of HRS and TLR7 in the spleens of EV71-infected TLR7 WT mice or TLR7-/- mice . As expected , TLR7 protein was only detected in the spleens of TLR7 WT mice ( S5A Fig , left ) , but not in the spleens of TLR7-/- mice ( S5A Fig , right ) . CD68 positive cells were not produced in the spleen of mock-infected TLR7 WT mice ( S5B Fig , left ) , but produced in the spleens of EV71-infected TLR7 WT mice ( S5B Fig , right ) , suggesting that inflammatory response is induced by EV71 infection in vivo [30] . Moreover , HRS co-localized with TLR7 in CD68 positive cells in the spleens of EV71-infected TLR7 WT mice ( Fig 6F ) . Taken together , we show that HRS binds and co-localizes with TLR7 in vitro and in vivo , and may act as a component in the TLR7 complex . The role of HRS in the regulation of the TLR7 complex was investigated by examining whether HRS is involved in dynamic endosomal events upon TLR7 activation . Endogenous clathrin and HRS did not co-localize in the cytoplasm of untreated macrophages , but most clathrin and HRS co-localized at foci in particular areas in the cytoplasm of R848-treated cells ( Fig 7A ) . Since clathrin participates in endosomal trafficking [31] , this result suggests that HRS may regulate TLR7 endosomal trafficking . TLRs must move from the ER to the Golgi to endosomes during intracellular stimulation [32 , 33] . We showed that TLR7 mainly co-localized with EEA1 and Rab7 ( endosome markers ) in R848-stimulated cells ( Fig 6E ) . Accordingly , when TLR7 signaling was activated by R848 , HRS mainly co-localized with EEA1 ( Fig 7B ) and Rab7 ( Fig 7C ) , but not with Rab11 ( a marker of recycling endosome ) ( Fig 7D ) , LAMP1 ( a marker of lysosome ) ( Fig 7E ) , Calnexin ( Fig 7F ) or Rcas1 ( Fig 7G ) ( a Golgi marker ) [29 , 34] in macrophages ( Fig 7H ) , these observations suggest that HRS locates to endosomes upon TLR7 activation . The subcellular compartments distinct from endocytic vesicles were then separated . In the absence of the TLR7 ligand R848 , TLR7 was present in fractions 3 to 7 as the ER manifested by Calnexin , whereas HRS was present in fractions 10 to 16 . In the presence of R848 , both TLR7 and HRS were present in fractions 10 to 16 , the early- and late- endosome marker proteins EEA1 and Rab7 were also present in fractions 10 to 16 ( Fig 8A ) , these findings suggest that HRS and TLR7 locate to the same components only upon TLR7 stimulation . The separated fractions were further analyzed by Co-IP as described previously [35] . TLR7 and HRS were detected only in the membrane fraction ( P100K ) , but not cytosol fraction ( S100K ) ( Fig 8B , lysate ) , and the interaction of TLR7 with HRS was enhanced by R848 stimulation ( Fig 8B , IP ) , suggesting that TLR7 and HRS located in the membrane upon TLR7 activation . In addition , HRS interacted with TLR7 , but HRS-dFYVE ( a mutant HRS with the FYVE domain deleted ) failed to interact ( Fig 8C ) , revealing that this domain is required for the interaction between TLR7 and HRS . Since FYVE domain is essential for the binding of HRS to the endosomal membrane [18] , our results suggest that HRS binds to TLR7 at the endosomal membrane . Moreover , NF-κB activity was stimulated by coexpression of TLR7 and HRS , but not by TLR7 with HRS-dFYVE ( Fig 8D ) , indicating that the FYVE domain of HRS is required for TLR7-mediated activation of NF-κB . Thus , HRS may recruit TLR7 in early- to late-endosomes to facilitate TLR7 activation , and the interaction between HRS and TLR7 may be important for TLR7 complex formation in endosomes . Induction of TLR7 signaling is initiated by MyD88 and IRAK binding that further recruits TRAF6 , TAK1 , and TAB1 to activate multiple intracellular signaling cascades , including MAPK , NF-κB , and IRF3/7 , leading to the production of proinflammatory cytokines [3 , 36] . Since proinflammatory cytokines are predominantly produced by macrophages and dendritic cells ( DCs ) [37] , we determined the role of HRS in the regulation of the TLR7 complex in macrophages . In R848-stimulated macrophages ( Fig 8E ) or EV71-infected macrophages ( Fig 8F ) , the interactions of TLR7 with MyD88 and IRAK1 were greatly reduced in cells transfected with siR-HRS . However , unexpectedly , the interaction between TLR7 and TAB1 and PKCζwas not reduced in HRS knock-down cells ( Fig 8E and 8F ) . These results suggest that HRS is important for the interaction of the TLR7 with MyD88 and IRAK1 . Because HRS is important for TLR7 activation , the effect of HRS on downstream events of TLR7 signaling was examined . p-p65 , p-IκBα , and p-p38 were reduced in siR-HRS transfected cells upon R848 stimulation ( Fig 9A ) and EV71 infection ( Fig 9B ) , revealing that HRS plays a critical role for TLR7-mediated activation of NF-κB and p38 . The role of HRS in the production of proinflammatory cytokines , a downstream event of the signaling pathway , was then evaluated . CSF3 , IL-1β , and IL-6 mRNAs and proteins were induced in R848-stimulated or EV71-infected macrophages , but this induction was reduced in the presence of siR-HRS ( Fig 9C and 9D ) , suggesting that HRS plays a critical role in the TLR7-induced production of proinflammatory cytokines . The effect of HRS on activation of TLR7 signaling was also verified in mouse primary cells . Mouse primary bone marrow-derived macrophages ( BMDMs ) were infected with lentivirus encoding shR-HRS ( Lenti-shHRS ) . HRS mRNA and protein were down-regulated significantly by Lenti-shHRS in BMDMs ( S6A Fig ) . IL-1β and IL-6 expression were induced by R848 , but this induction was reduced in cells transfected with Lenti-shHRS ( S6B Fig ) . Intracellular cytokine staining showed that IL-6 protein expression was induced by R848 , but not in the presence of Lenti-shHRS in CD14+ cells ( S6C Fig ) , suggesting that HRS is important for TLR7-mediated activation of proinflammatory cytokines in mouse primary cells during the signaling initiation . To confirm the function of HRS in TLR7-triggered inflammatory responses , we mimicked inflammation in vivo . Numbers of CD68+ cells were significantly increased in the spleen of mice treated with R848 ( S6D Fig ) , indicating that an inflammatory response is induced in macrophages during maturation and activation as demonstrated previously [30] . Mice were tail vein injected with siR-HRS-m ( a 5’-cholesterol-modified siRNA duplex targeted to HRS ) and siR-NC-m ( its control ) , followed by intraperitoneal injection with R848 , and then PBMCs and splenocytes were isolated ( Fig 9E ) . HRS and IL-6 mRNA were induced by R848 , but significantly reduced in the presence of siR-HRS-m in PBMCs and splenocytes ( Fig 9F and 9G ) . Moreover , IL-1β , IL-6 , TNFα and IL-10 protein expression was induced after R848 stimulation , but this induction was reduced by siR-HRS-m in sera of mice ( Fig 9H ) . These results suggest that knock-down of HRS down-regulates TLR7-mediated production of proinflammatory cytokine . Taken together , HRS plays a vital role in proinflammatory cytokine production during TLR7-mediated inflammation in vitro and in vivo upon R848 stimulation . The role of HRS in the regulation of IFN signaling in response to virus infection was investigated . In R848-stimulated macrophages , phosphorylation of IRF3 was reduced in cells transfected with siR-HRS ( Fig 10A ) . Similarly , in EV71-infected macrophages , phosphorylation of IRF3 was reduced by siR-HRS ( Fig 10B , left vs . right ) . IFN-β mRNA expression was induced by SeV , vesicular stomatitis virus ( VSV ) and influenza A virus ( IAV ) infection , whereas the induction was abrogated in the presence of siR-HRS ( Fig 10C , upper ) ; IFN-λ1 mRNA expression was activated by EV71 , SeV , VSV , and IAV infection , whereas the induction was reduced by siR-HRS ( Fig 10C , bottom ) . We also noticed that IFN-β mRNA was not induced by EV71 or HCV infection , and IFN-λ1 mRNA was not induced by HCV infection . In addition , IFN-β was induced by SeV and IAV infection and it was further enhanced by HRS overexpression ( Fig 10D , upper ) ; while IFN-λ1 was induced by EV71 , SeV , and IAV infection and further facilitated by HRS overexpression ( Fig 10D , bottom ) . These results demonstrate that HRS promotes TLR7-mediated IFN production during viral infection . Moreover , IFN-β mRNA expression was induced by R848 and this was reduced by Lenti-shHRS in mouse BMDMs ( S7A Fig ) , and IFN-β protein expression was enhanced by R848 and this enhancement was reduced by Lenti-shHRS in CD14+ cells ( S7B Fig ) , indicating that HRS plays a role in TLR7-mediated activation of IFN-β in mouse primary cells . Furthermore , HRS and IFN-β mRNAs expression was enhanced by R848 , but reduced in the presence of siR-HRS-m in mouse PBMCs ( Fig 10E ) and splenocytes ( Fig 10F ) . IFN-α and IFN-β protein expression was induced by R848 , but repressed in the presence of siR-HRS-m in mouse sera ( Fig 10G ) . In conclusion , HRS plays an important role for TLR7 mediated IFN production in human macrophages and in mice .
Activation of TLRs initiates the integration of contextual cues and signals to regulate host inflammatory and immune responses [38] . TLR7 recognizes viral ssRNA to induce a wide range of proinflammatory cytokines and IFNs [39] . Here , we reveal a novel mechanism underlying the regulation of TLR7 signaling . Clinical investigation , animal study , and cellular analysis demonstrate that the productions of 3 important proinflammatory cytokines CSF3 , IL-1β , and IL-6 are induced during EV71 infection . TLR7/NF-κB and TLR7/MAPK pathways are involved in such activation , which is consistent with a previous report showing that TLR7 activates MAPK and NF-κB signaling [40] . EV71 induces multiple TLRs accompanied with excessive production inflammatory factors . In intestinal epithelial cell , TLR7 mRNA is upregulated by EV71 , suggesting that TLR7 mediates inflammatory response during EV71 infection [14 , 15] . EV71 infection also stimulates TLR3 signaling in macrophage , leading to activation of iNKT ( invariant natural killer T ) cells [25] . In our study , we illustrate that EV71 infection facilitates TLR7 signaling in macrophages , resulting in induction of proinflammatory cytokines . By using a stable cell line HEK293T/TLR7/NF-κB , we discovered that TLR7 , MyD88 , and IRAK1 are required for EV71-induced activation of NF-κB . Mini-library RNAi screening revealed that HRS is involved in TLR7/NF-κB-mediated inflammatory response . TLR7 and HRS are induced and highly correlated in EV71-infected patients , mice and cultured cells , which implicates that TLR7 and HRS shared similar function or had protein-protein interaction based on a co-expression prediction [41] . In the process of identifying HRS-interacting proteins , we reveal that HRS binds directly with TLR7 and TAB1 , and co-localizes with TLR7 and TAB1 at foci in specific areas in the cytoplasm , indicating that HRS acts as a regulator or component of the TLR7 complex . HRS mediates endosome trafficking by binding with targeted proteins to deliver internalized cargos in the endocytic pathway [42] . We showed that HRS and TLR7 co-localize with clathrin at foci in the cytoplasm of infected cells . Because clathrin mediates endosomal trafficking events [31] and endosomes are the major sites for TLR complex assembly and antiviral response initiation [33 , 43] , we suggest that HRS regulates TLR7 signaling in the endosomes . Upon stimulation , intracellular TLR7 moves from the ER to endosome via a clathrin-dependent endocytic pathway [44] , a series of accessory molecules act as delivery cofactors , chaperones or trafficking proteins for biosynthesis and activation of TLR7 [45 , 46] . Our results indicate that HRS is a key cofactor for TLR7 activation in endosomal location , in which HRS recruits TLR7 to facilitate it from early-endosome to late-endosome . We showed that the FYVE domain is required for HRS in the facilitation of TLR7 complex assembly from early- to late-endosome , which confirms that HRS enhances TLR7 signaling in the endosome . Thus , HRS regulates the TLR7 complex components in endosomes and may play a role in the initiation of immune response . In bone morphogenetic protein ( BMP ) signaling , HRS facilitates the crosstalk between SMADs and TAK1 , which subsequently stabilizes the interaction between Smad2/Smad3 and TGF-β receptor and functions in internalization with the endocytic machinery in endosomes [47 , 48] . We demonstrate that HRS acts as a key component in TLR7 signaling to facilitate the assembly of the TLR7/MyD88/IRAK1 complex by recruiting TLR7 , and plays important roles in TLR7 signaling activation and initiation of the inflammatory response . TLR7 stimulates signaling through TIR domain , triggers the binding of MyD88 , activates IRAK1 , and recruits TRAF6 and TBK1 [49] . Recruitment of the cellular factors activates multiple intracellular signaling cascades MAPK , NF-κB , and IRF3/7 to induce proinflammatory cytokines and IFNs [3] . We reveal that HRS is important for activation of NF-κB , IκBα , and p38 during viral infections , suggesting that HRS plays critical roles in TLR7-mediated activation of NF-κB , p38 MAPK , and IRF3 pathways . TLR7 in the endosomes and endolysosomes plays an important role in the initiation of antiviral responses [43] . We demonstrate that HRS is involved in TLR7-mediated induction of IL-1β and IL-6 in mice primary BMDMs and human cells , and stimulates cytokine production by facilitating the TLR7 complex assembly and NF-κB signaling . In summary , we reveal a novel mechanism underlying the regulation of TLR7 signaling ( Fig 11 ) . EV71 infection initially induces the production of HRS , which subsequently acts as a key regulator or component of the TLR7 complex coupling with endosomal location by binding with TLR7 and TAB1 to facilitate the assembly of the TLR7 complex . On one hand , HRS activates TLR7/NF-κB/IRF3 signaling to promote the production of IFNs , resulting in induction of innate immune response . On the other hand , HRS simulates TLR7/NF-κB/p38 signaling to enhance production of proinflammatory cytokines , leading to the induction of inflammatory response . Therefore , HRS-mediated TLR7 complex assembly may provide an important mechanism for the regulation of host immune and inflammatory responses during viral infection .
The cases of mild HFMD were defined as patients with vesicular lesions on their palms , feet , and mouth , with or without fever , and severe HFMD were accompanied by neurologic or cardiopulmonary complications . All participants were diagnosed with EV71 infection by presence of EV71 RNA with specific EV71 VP1 primer and confirmed as negative for other enterovirus . Peripheral blood specimens were obtained from 40 EV71-infected patients hospitalized at the Wuhan Medical Treatment Center and Wuhan Infectious Diseases Hospital from May 2 , 2016 to May 10 , 2017 . The peripheral blood samples were obtained from all subjects on admission from 12 to 24 h . The blood samples were immersed in ice and transported immediately to the laboratory for processing . Matched by age and sex , peripheral blood specimens were collected from 36 EV71-negative healthy individuals in a local blood donation center as controls . All individuals did not suffer any concomitant disease at the moment of sampling , did not show any serological markers suggestive of autoimmune disease , and had not received any antiviral or immunomodulatory therapy prior to this study . BALB/c mice were purchased from Shanghai Laboratory Animal Center . The mice were housed under specific pathogen-free conditions in individually ventilated cages . One-day-old suckling mice were intraperitoneally ( i . p . ) mock-infected or infected with 1×107 plaque-forming units ( PFU ) of MA-EV71 in 50 μl PBS . Following EV71 infection , the mice were scored as follows: 0 , healthy; 1 , ruffled hair and hunchbacked; 2 , limb weakness; 3 , paralysis in one limb; 4 , paralysis in both limbs; and 5 , death . For EV71 infection mice model , mice were sacrificed at 10 days post-infection . In siRNA in vivo transfection experiments , 5’-cholesterol-modified siRNA duplexes ( siR-NC: 5’-TTCTCCGAACGTGTCACGT; siR-HRS-mouse: 5’-CGCAUGAAGAGCAACCACA; GenePharma , Shanghai , China ) were diluted in 0 . 9% NaCl solution and injected into a mouse ( weight = 20 g ) via the tail vein ( injection volume = 100 μl ) based on a final dose of 2 mg/kg twice for 2 d . Subsequently , R848 ( a TLR7 agonist ) diluted in PBS was injected intraperitoneally into a mouse at a final dose of 2 mg/kg . Two days after the final injection , mice were sacrificed and spleen cells and PBMCs were isolated for analyses . The spleen tissues were fixed in 3 . 7% paraformaldehyde and then applied to immunohistochemistry ( IHC ) or immunofluorescence ( IF ) staining . The study was conducted according to the principles of the Declaration of Helsinki and approved by the Institutional Review Board of the College of Life Sciences , Wuhan University , in accordance with its guidelines for the protection of human subjects . All participants have provided written informed consent to participate in the study . All animal studies were performed in accordance with the principles described by the Animal Welfare Act and the National Institutes of Health guidelines for the care and use of laboratory animals in biomedical research . All procedures involving mice and experimental protocols were approved by the Institutional Animal Care and Use Committee ( IACUC ) of the College of Life Sciences , Wuhan University ( Permit numbers: 2015–006 ) . The growth , virus titration and inoculation of enterovirus 71 ( EV71 ) [50 , 51] , Sendai virus ( SeV ) [52] , vesicular stomatitis virus ( VSV ) , influenza A virus ( IAV ) [53] , and hepatitis C virus ( HCV ) [54] were performed as described previously . EV71 virus strain ( Xiangyang-Hubei-09 ) was isolated previously in our laboratory ( GenBank accession no . JN230523 . 1 ) . Virus stock was propagated in RD cells . Indiana serotypes of VSV strain were provided by the China Center Type Culture Collection . IAV strain A/HongKong/498/97 ( H3N2 ) was provided by the China Center for Type Culture Collection . The virus stock was propagated in human lung epithelial cells ( A549 ) were cultured in F12K medium ( Invitrogen ) . HCV genotype 2a strain JFH-1 was kindly provided by Takaji Wakita . Huh7 . 5 . 1 cells were infected with JFH-1 at a multiplicity of infection ( MOI ) of between 0 . 1 and 5 . HCV was propagated for 6 days before collection . Virus stock was obtained after filtering of the cell supernatant . Viral titers were quantified using a commercial kit ( HCV RNA qPCR diagnostic kit; KHB Company , Shanghai , China ) . Aliquots were stored at -80°C prior to use . Cells were infected with virus at the indicated multiplicities of infection ( MOI ) and unbound virus was washed away 2 h later , and then incubated at 37°C for an additional 10 h . Human leukemic monocytes ( THP-1 ) cells , human embryonic kidney 293T ( HEK 293T ) cells , human lung adenocarcinoma A549 ( A549 ) cells , human hepatoma ( Huh7 . 5 . 1 ) cells , and mouse Raw264 . 7 cells were purchased from American Type Culture Collection ( ATCC ) ( Manassas , VA , USA ) . Human embryonal rhabdomyosarcoma ( RD ) cells and HeLa cells were obtained from China Center for Type Culture Collection ( CCTCC ) ( Wuhan , China ) . THP-1 cells were cultured in RPMI 1640 medium ( Invitrogen , Carlsbad , CA ) , supplemented with 10% fetal bovine serum ( FBS ) , 100 U/ml penicillin and 100 mg/ml streptomycin sulfate at 37°C in a 5% CO2 incubator . Macrophages were derived from THP-1 cells induced by 100 nM 12-O-tetradecanoyl-phorbol-13-acetate ( TPA ) ( Sigma , St . Louis , MO ) for 72 h . Cell surface markers for differentiated macrophages were detected by flow cytometry . PBMCs were obtained by density centrifugation of peripheral blood samples diluted 1:1 in pyrogen-free saline over Histopaque ( Haoyang Biotech , Beijing , China ) . Cells were washed twice in saline and suspended in culture medium . HEK293T/TLR7/NF-κB stable cell lines were generated after co-transfecting HEK293T cells with TLR7 expression plasmid pTLR7 and 5 × NF-κB luciferase reporter construct . For the culture of mouse BMDMs , male C57BL/6 TLR7-/- mice ( a kind gift from Dr . Ling Zhao , Huazhong Agricultural University , Wuhan , China ) 4–6 weeks of age were sacrificed and pelvic and femoral bones collected to separate bone marrow . Cells were cultured in RPMI 1640 medium with 10% fetal bovine serum ( FBS ) following induction with 100 ng/ml mouse granulocyte-macrophage colony-stimulating factor ( GM-CSF ) ( Peprotech , NJ , USA ) for 3 d until cells became attached BMDMs . Rabbit antibodies against TLR7 , MyD88 , NF-κB p65 , and phosphorylated IκBα ( p-IκBα , Ser32 ) were purchased from Abcam ( Cambridge , United Kingdom ) . Mouse antibodies against HRS , IRAK1 , and c-Myc Tag , goat antibodies to TAB1 , and normal rabbit IgG were purchased from Santa Cruz Biotechnology ( Santa Cruz , CA ) . Rabbit antibodies against phosphorylated NF-κB p65 ( p-p65 , Ser536 ) , phosphorylated p38 MAPK ( p-p38 , Thr180/Tyr182 ) , p38 MAPK , Clathrin , Calnexin , EEA1 , Rab5 , Rab7 , Rab11 , LAMP1 , Rcas1 , and TRAF6 were purchased from Cell Signaling Technology ( Beverly , MA ) . Rabbit antibodies against GFP , PKCζ , β-actin , and GAPDH were purchased from ProteinTech Group ( Chicago , IL ) . Mouse antibody to EV71 VP1 was from Abnova Company ( Taipei ) . Rabbit antibody to EV71 3C was raised against residues 76–88 of 3C protein ( Abgent , Suzhou , China ) . Rat antibodies to FLAG , and anti-mouse CD68 were purchased from BioLegend ( San Diego , CA ) . The specific small interfering RNA ( siRNA ) to the negative control ( NC ) , HRS , and mini-siRNA library were synthesized by RiboBio ( Guangzhou , China ) . siR-NC targeting at sequence: 5’-TTCTCCGAACGTGTCACGT; siR-HRS targeting at sequence: #1: 5’-AGGTAAACGTCCGTAACAA; #2: 5’-GCATGAAGAGTAACCACAT . All kinase inhibitors were purchased from Sigma Chemical Company ( St . Louis , MO ) and dissolved in DMSO upon use as described previously [54] . R848 ( InvivoGen ) dissolved in ddH2O and used at a final concentration of 100 ng/ml . The full-length TLR7 gene ( GenBank accession no . NM_016562 ) was amplified from THP-1 cells and the cDNA was inserted into the pCMV-Flag2B vector ( Invitrogen ) to create pFlag-TLR7 , which encoded a Flag-TLR7 fusion protein . The plasmid pFlag-TLR7 ( Y892A ) expressing a mutant Flag-TLR7 ( Y892A ) fusion protein was constructed as described previously [21] . The full-length and truncated HRS genes ( GenBank accession no . NM_004712 ) were cloned into pcDNA3 . 1 ( - ) -Myc-His vector ( Invitrogen ) to generate plasmids encoding for HRS-Myc and dFYVE-HRS-Myc ( with a deletion of amino acids 166–215 ) . The DNA fragments of TLR7 , MyD88 , IRAK1 , TRAF3 , TRAF6 , TAB1 , TAK1 , and TBK1 genes were ligated into eGFP-C1 ( Takara ) vector to generate plasmids expressing GFP-tagged TLR7 , MyD88 , IRAK1 , TRAF3 , TRAF6 , TAB1 , TAK1 , and TBK1 proteins , respectively . The shTLR7 plasmid expressing a short-hairpin RNA ( shRNA ) ( 5’-GCCAACAACCGGCTTGATTTA ) targeting TLR7 was generated by inserting the DNA fragment into the pSilencer 2 . 1-U6 neo vector ( Ambion , Inc . , Austin , TX ) . shGFP , shMyD88 , shIRAK1 , shp65 , and shp50 were constructed as described previously according to targeting sequences of GFP , MyD88 , IRAK1 , p65 , and p50 , respectively [52] . The primers used in this study are listed in S3 Table . HEK 293T cells ( 1×105 cells ) were co-transfected with pFlag-TLR7 or vector ( 0 . 24 μg ) and NF-κB luciferase reporter ( 0 . 08 μg ) using Lipofectamine 2000 ( Invitrogen ) in each well of 24-well plate . The determination of NF-κB reporter luciferase activity was performed as described previously [54] . HEK 293T cells ( 1×105 cells ) were co-transfected with the luciferase reporter plasmid ( 80 ng ) , pRL-TK Renilla luciferase control plasmid ( 40 ng ) , and the indicated siRNA ( final concentration , 20 nM ) . After 24 h , luciferase activities were determined with the Dual-Luciferase Reporter Assay System ( Promega ) according to the manufacturer’s instructions . Data were normalized for transfection efficiency by dividing firefly luciferase activity with that of Renilla luciferase . For knock-down HRS , the specific HRS siRNA or control siRNA were transfected into THP-1 cells using INTERFERin ( Polyplus Transfection Company , Illkirch , France ) . Cells were treated with R848 or EV71 at 36 h post-transfection , and total cellular RNA was extracted using TRIzol reagent ( Invitrogen ) . Reverse transcription and quantitative PCR ( qPCR ) was performed as described previously [54] . Quantitative RT-PCR ( qPCR ) analysis was performed using the Roche LightCycler 480 and SYBR RT-PCR kits ( Roche ) ; each 20 μl reaction contained 0 . 5 μM of each PCR primer , 10 μl of SYBR Green PCR master mix , 1 μl of diluted template , and RNase-free water . The data represent absolute mRNA copy numbers normalized to GAPDH used as a reference gene . Relative fold expression values were determined by using the ΔΔCt method . The information of real-time PCR primers is listed in Table 2 . The tested proteins in the sera of clinical samples were measured with Human IL-1β immunoassay ( BD Biosciences , CA ) and Human CSF3 and IL-6 immunoassay ( R & D systems , Minneapolis , MN ) . Proteins in mice sera were measured with Mouse CSF3 , IL-6 , IL-1β , IL-10 and TNFα immunoassay ( 4A Biotech , Beijing , China ) , as well as IFN-α and IFN-β immunoassay ( eBioscience; Santa Diego , CA ) following the manufacturer’s instructions , respectively . The 30 cytokines ELISA assay for EV71-infected THP-1 supernatant was described in details as follow . THP-1 cells ( 1×106 per well ) were distributed into 6-well culture plates with 10% FBS , and then incubation with EV71 at an MOI = 5 for 2 h at 37°C . Suspended cells were centrifuged at 1000 g for 10 min . Cells were resuspended in PBS to wash away the unbound virus . After three times washing in PBS , cells were cultured in 2 ml of FBS-free RPMI 1640 for additional 12 h . Finally , the cell supernatant was harvested after the centrifugation at 10000 g for 5 min for ELISA assay . 30 cytokines in cell supernatants were measured with indicated antibodies-coated plate of Human Cytokine ELISA Plate Assay I kit ( Signosis , Sunnyvale , CA , USA ) . The simultaneous quantification of each cytokines concentration in cell supernatants was determined according to absorbance value . Each cytokine induction was normalized to mock control and data are shown as fold changes in a heatmap visualization . Human THP-1 cells ( 5×106 cells ) were cultured in 10 cm dish and harvested after R848 stimulation as indicated time points and lysed in 1 ml RIPA buffer ( 50 mmol/l Tris-HCl pH7 . 4 , 150 mmol/l NaCl , 1% sodium deoxycholate , 0 . 5 mol/l EDTA , 1 mmol/l NaF , 1% Nonidet P-40 , supplemented by 10% proteinase inhibitors cocktail ) . 1/10 lysate ( 100 μl ) was reserved for direct immunoblot analysis while the rest was successively incubated with rabbit anti-TLR7 or normal rabbit IgG antibodies for 4 h and with protein G agarose ( GE Healthcare ) for another 1 h . After 5 rounds of washes , proteins were fractionated by SDS-PAGE and transferred to nitrocellulose membrane ( Amersham , Piscataway , NY ) . Nonspecific sites were blocked with 5% nonfat dried milk ( BD Biosciences , San Jose , CA ) for 1 h at room temperature . After 3 times of PBS wash , the nitrocellulose membrane was incubated with primary and secondary antibodies . Blots were analyzed using a Luminescent Image Analyzer ( Fujifilm LAS-4000 ) . The expression and purification of GST or GST-HRS proteins were performed as follows . The cDNA encoding HRS was cloned into pGEX6p-1 vector ( GE Healthcare ) and transformed to E . coli BL21 strain . A single colony was grown in 5 ml LB culture containing 100 μg/ml of Ampicillin overnight . 1 ml of bacteria culture was then transferred to a 50 ml flask with 20 ml of LB and 100 μg/ml of Ampicillin in an orbital shaker at 37°C , 200 rpm for 2 h until A600 = 0 . 5–0 . 6 . IPTG was incubated with the culture at a final concentration of 0 . 6 mM at 37°C , 200 rpm for additional 6 h . Cells were centrifuged at 8 , 000 g for 10 min at 4°C and the supernatant was removed . The pellets were resuspended with 10 ml PBS containing 0 . 1 mM PMSF and lysed with the ultrasonic processor . The lysate was centrifuged at 8 , 000 g for 10 min at 4°C and the supernatant was prepared . Before the affinity of the GST-fusion to the column , 2 ml of glutathione-sepharose beads ( GE Healthcare ) was equilibrated with 10 volume of lysis buffer ( 50 mM Tris-HCl pH8 . 0 , 150 mM NaCl , 1 mM DTT , 0 . 1 mM EDTA and 0 . 1 mM PMSF ) at 4°C . Unbound lysis was washed away with 15 ml wash buffer ( 50 mM Tris-HCl pH8 . 0 , 150 mM NaCl , 1 mM DTT ) . Finally , GST-fusion protein was eluted with 5 ml of elution buffer ( 125 mM Tris-HCl pH8 . 0 , 45 mM NaCl , 0 . 0001% Triton X-100 , 5 mM EDTA and 30 mM reduced Glutathione ) . The cell lysate was incubated with 50 μl of glutathione-sepharose beads ( 50/50 slurry in lysis buffer ) and 25 μg of GST or GST-HRS for 2 h at 4°C with end-over-end mixing . The beads were washed four times with 1 ml of ice-cold lysis buffer . After discarding the last wash , proteins bound to the probe protein were dissociated for immunoblot analysis . Immunofluorescence microscopy experiments were performed as described previously [50] . Briefly , macrophages were seeded on 20-mm cover slips and then treated with R848 for 30 min , or infected with EV71 or SeV for 4 h . Then , the culture medium was removed and cells were washed with PBS , fixed with 3 . 7% formaldehyde for 20 min , and permeabilized using 0 . 4% Triton X-100 for 5 min at room temperature . After another PBS wash , the cells were blocked in PBS containing 5% BSA for 1 h , incubated with antibodies for 3 h at 37°C . The samples were incubated with FITC-conjugated donkey anti-rabbit immunoglobulin G ( IgG ) and Cy3-conjugated donkey anti-mouse IgG ( ProteinTech Group ) for 45 min at room temperature . To stain nuclei , 1 mg/ml DAPI ( Roche ) methanol solution was added and samples incubated for 15 min at room temperature . After washing with PBS , samples were visualized by confocal laser scanning microscopy ( Fluoview FV1000; Olympus , Tokyo , Japan ) . The N-terminal truncated ( VN-173 ) and the C-terminal truncated ( VC-155 ) version of nonfluorescent Venus YFP fragments vectors were brought here ( Addgene plasmid # 22010 ) [55] . Human TLR7 and TAB1 cDNA were subcloned in VN-173 , and human HRS cDNA was ligated to VC-155 without their stop codons using ClonExpress MultiS One Step Cloning Kit ( Vazyme Biotech Co . , Ltd , Nanjing , China ) , respectively . The resulting plasmids or empty vectors were cotransfected into HEK293T cells using Lipofectamine 2000 . At 24 h post-transfection , cells were pre-cultured at 4°C for 10 min . The fusion proteins in living cells were observed by confocal microscopy . Macrophages were treated with or without R848 ( 100 ng/ml ) for 30 min . The harvested cells were washed extensively with PBS . The pellet obtained following centrifugation of cells for 5 min at 2 , 000 g was washed twice with TS buffer ( 0 . 25 M sucrose , 10 mM Tris-Cl , pH 7 . 4 ) then were homogenized with 40 strokes in a Douncer on ice , followed by centrifugation at 1 , 000 g for 20 min at 4°C . To separate the membrane pellets from the cytosol , the supernatant fractions were subjected to further centrifugation at 100 , 000 g . To separate subcellular fractionation , the supernatant was adjusted by adding 1 . 2 volume of 62% sucrose , sequentially overlaid with 1 . 5 volume of 35% sucrose , 1 volume of 25% sucrose and filled up with TS buffer to the rest of the tube . Sucrose gradient centrifugation was performed in a Beckman SW41 Ti rotor at 210 , 000 g at 4°C for 2 h . The fractions were collected from the top to bottom by subsequent detection in the distribution of intracellular markers by SDS-PAGE and Western blotting . Recombinant lentivirus carrying shRNA targeting HRS ( Lenti-shHRS-1: 5'-AAAGGTAAACGTCCGTAACAA , and Lenti-shHRS-2: 5’- CCGCATGAAGAGTAACCACAT ) or the control ( Lenti-shNC: 5’-CAACAAGATGAAGAGCACCAA ) were constructed from plasmids psPAX2 , pMD2 . G and pLKO . 1 ( Addgene , Cambridge , MA ) and generated from co-transfected HEK293T cells . Cells were seeded in 96-well plate after treatment . Cell viability was determined by CellTiter 96 AQueous One Solution Cell Proliferation Assay ( Promega ) according to the instructions provided by the manufacturer . Flow cytometric measurements were performed using a Beckman Coulter flow cytometry ( Fullerton , CA ) . For detection of cell surface markers , Fc receptors were blocked by incubating 100 mg recombinant human IgG ( Sigma-Aldrich ) with cells for 15 min at 4°C prior to antibody staining . 1 mg of monoclonal PE mouse IgG1κ anti-human CD14 antibody , or the relevant PE mouse IgG1κ Isotype Ctrl ( FC ) antibody ( BioLegend ) was incubated with samples containing 2 × 105 cells for 15 min at 4°C . Following incubation samples were washed and resuspended in phosphate buffered saline ( PBS ) and 10 , 000 events recorded . Data was analyzed using Summit software , version 5 . 0 . 1 ( Beckman Coulter Inc . ) . PE anti-human CD14 or isotype , PE/Cy7 anti-mouse CD14 or isotype control , PE anti-mouse IL-6 or isotype control were purchased from BioLegend ( San Diego , CA ) . FITC anti-mouse IFN-β or isotype control antibodies were from R&D systems ( Minneapolis , MN ) . For intracellular cytokine staining , BD GolgiPlus™ ( 1 μg/ml ) and BD GolgiStop ( 0 . 7 μg/ml ) protein transport inhibitor ( BD Biosciences , CA ) was added for 5 h during antigen stimulation . Intracellular cytokines were stained with indicated antibodies with the Cytofix/Cytoperm Kit ( BD Biosciences ) . Cells were acquired on a FACS Calibur flow cytometer ( BD Biosciences ) , and data were analyzed with FlowJo software ( Tree Star , Inc , Ashland , OR ) . All experiments were reproducible and each set was repeated at least three times with similar results . All data were recorded as means ± standard deviation ( SD ) unless stated otherwise . Statistical testing was performed using Prism 5 software ( GraphPad Software Inc . ) with the statistical tests indicated in the figure legends . A P-value < 0 . 05 was considered to indicate statistical significance . | Enterovirus 71 ( EV71 ) is a highly infectious positive-stranded RNA virus that causes hand-foot-mouth disease ( HFMD ) . As a major pathogen , EV71 infection leads to host immune responses in the disease severity . Toll-like receptors ( TLRs ) can recognize pathogens to induce host immunity and inflammation . Most TLRs must traffic from the endoplasmic reticulum ( ER ) to endolysosomal network before responding to ligands . The hepatocyte growth factor-regulated tyrosine kinase substrate ( HRS ) regulates ESCRT-0 complex and endosomal sorting of membrane proteins . HRS is required for ubiquitin-dependent TLR9 targeting to the endolysosome , however , the mechanism by which HRS regulates inflammation and immunity mediated by TLR7 is still largely unknown . Here , we reveal that HRS is a key component of TLR7 signaling to orchestrate immunity and inflammation during EV71 infection . EV71 infection induces the expression of HRS , which subsequently enhances the TLR7 complex formation by binding with TLR7 and TAB1 . HRS facilitates TLR7/NF-κB/p38 MAPK and TLR7/NF-κB/IRF3 signaling pathways to produce proinflammatory cytokines and interferons , leading to induction of inflammatory and immune responses . Thus , we identify HRS as a key regulator of TLR7 signaling and illustrate a novel mechanism underlying the regulation of host immunity and inflammation during viral infection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
]
| [
"blood",
"cells",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"cytokines",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"enzyme-linked",
"immunoassays",
"gene",
"regulation",
"immunology",
"immune",
"receptor",
"signaling",
"developmental",
"biology",
"signs",
"and",
"symptoms",
"membrane",
"receptor",
"signaling",
"molecular",
"development",
"immunologic",
"techniques",
"mapk",
"signaling",
"cascades",
"research",
"and",
"analysis",
"methods",
"immune",
"system",
"proteins",
"small",
"interfering",
"rnas",
"white",
"blood",
"cells",
"inflammation",
"animal",
"cells",
"proteins",
"gene",
"expression",
"immunoassays",
"immune",
"response",
"immune",
"system",
"toll-like",
"receptors",
"biochemistry",
"signal",
"transduction",
"rna",
"diagnostic",
"medicine",
"cell",
"biology",
"nucleic",
"acids",
"physiology",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"immune",
"receptors",
"macrophages",
"non-coding",
"rna",
"cell",
"signaling",
"signaling",
"cascades"
]
| 2017 | HRS plays an important role for TLR7 signaling to orchestrate inflammation and innate immunity upon EV71 infection |
During embryonic development , vascular networks remodel to meet the increasing demand of growing tissues for oxygen and nutrients . This is achieved by the pruning of redundant blood vessel segments , which then allows more efficient blood flow patterns . Because of the lack of an in vivo system suitable for high-resolution live imaging , the dynamics of the pruning process have not been described in detail . Here , we present the subintestinal vein ( SIV ) plexus of the zebrafish embryo as a novel model to study pruning at the cellular level . We show that blood vessel regression is a coordinated process of cell rearrangements involving lumen collapse and cell–cell contact resolution . Interestingly , the cellular rearrangements during pruning resemble endothelial cell behavior during vessel fusion in a reversed order . In pruning segments , endothelial cells first migrate toward opposing sides where they join the parental vascular branches , thus remodeling the multicellular segment into a unicellular connection . Often , the lumen is maintained throughout this process , and transient unicellular tubes form through cell self-fusion . In a second step , the unicellular connection is resolved unilaterally , and the pruning cell rejoins the opposing branch . Thus , we show for the first time that various cellular activities are coordinated to achieve blood vessel pruning and define two different morphogenetic pathways , which are selected by the flow environment .
The vascular system of vertebrates distributes oxygen , nutrients , metabolites , and blood cells to and from all organs of the body . A complex network of interconnected vascular tubes develops and already functions at early stages of embryonic development and continues to expand and to remodel as the animal grows [1] . Blood vessels invade avascular tissue areas by sprouting angiogenesis , which is defined by the branching of new vessels from existing ones [2 , 3] . Later on , angiogenic sprouts connect to each other in a process termed vascular anastomosis [4 , 5] . Emerging vascular beds are often organized in a plexus , which , in a primitive state , constitutes a nonhierarchical network of blood vessels . As the demand for oxygen and nutrients during organ or embryonic growth increases , these primitive networks become remodeled to allow more efficient blood transport . Such vascular remodeling involves changes in vessel diameter as well as pruning of supernumerary vascular branches [4] . The pruning process has been studied in the vasculature of the mouse retina [6 , 7] and extra-embryonic vessels of chick and mouse embryos [8–10] and , more recently , in several vascular beds of the zebrafish embryo [11 , 12] . Taking advantage of the ease of performing live imaging in the zebrafish embryo , the latter studies showed that regression of vessels occurs through endothelial cell rearrangements typically in the absence of apoptosis . Importantly , these studies showed that pruning is regulated by hemodynamic forces , since blood vessel regression was preceded by changes in blood flow patterns due to rewiring of the vascular plexus . We have previously shown that blood vessel fusion occurs in discrete morphogenetic steps , in which cell rearrangements play a crucial role in the transformation of blood vessel architecture from a unicellular to multicellular configuration [13 , 14] . The studies by Kochhan et al . [12] pointed out that cell activities involved in vessel pruning might resemble those during vessel fusion but in reversed order . However , the anatomical localization of the vessels analyzed in that report made it impossible to describe the pruning process at high cellular resolution . To analyze the pruning process in more detail , we have now characterized and established the subintestinal vein ( SIV ) plexus as a new model for vascular pruning . The SIV is a blood vessel network that delivers blood to the digestive tract of the zebrafish larva . The SIV grows over the yolk sack , ventral to the posterior cardinal vein ( PCV ) and posterior to the common cardinal vein ( CCV ) [15] . Here , we apply high-resolution time-lapse microscopy techniques and several transgenic reporter lines to analyze the formation and the remodeling of the SIV plexus in developing zebrafish embryos . We show that different cellular processes such as single cell sprouting , vessel fusion , and vessel pruning are involved and timely coordinated to give the plexus its final shape . We describe in detail the cellular activities underlying different modes of vessel pruning , which are achieved through a variety of cell rearrangements and cell shape changes . The choreography of these processes during pruning resembles vessel fusion “in reverse” and can follow two different pathways determined by the flow environment in the pruning branch .
The plexus of the SIV emerges from the PCV ~2 days post-fertilization ( dpf ) as a reticular structure , which spreads bilaterally over the surface of the yolk [15] . By ~5 dpf , it has formed a set of parallel vessels situated along the dorsoventral axis and connected on the dorsal side to the subintestinal artery ( SIA ) ; a large vessel on the ventral side of the SIV plexus collects blood from all the parallel SIV branches and brings it back towards the heart . Because a detailed description of SIV formation is missing , we followed the development of the entire plexus over several days by using Selective Plane Illumination Microscopy ( SPIM ) , a novel technique for fast , long-term imaging of large fields of view [16] . In SPIM , images from different angles are acquired by rotating the sample , which is particularly beneficial for the large and curved structure of the SIV . Furthermore , the noninvasive embryo mounting [17] combined with low phototoxicity minimize interference of the imaging with the development of the vasculature [18] . We imaged Tg ( fli1a:EGFP ) y1 transgenic zebrafish embryos over 48 h ( from ~36 to ~84 hours post-fertilization [hpf] ) ( S1 Movie and S2 Movie and Fig 1A–1E’ ) . Based on these observations , we can divide the development of the SIV into four phases: ( I ) formation of the primary SIV tube through ventral sprouting of endothelial cells from the PCV followed by cell coalescence ( Fig 1A–1A’ , see S1A Fig and S3 Movie for details ) ; ( II ) vascular loop formation through fusion of angiogenic sprouts originating from the primary SIV ( Fig 1B–1B’ , see S1B Fig and S4 Movie for details ) ; ( III ) formation of a reticular structure with multiple vascular loops ( Fig 1C–1C’ ) ; and ( IV ) remodeling into the final structure with parallel , vertical branches that drain into the large ventral SIV ( Fig 1D–1E’ ) . The transition from phase III to IV involves extensive remodeling of the reticular structure , leading to a reduction of the number of loops and a redirection of the flow to the major vertical branches . Elimination of vascular loops occurs through regression of supernumerary cross branches or by collateral fusion of a cross branch to a neighboring major branch ( S2 Fig and S5 Movie ) . Because of variability in the sprouting phase , the number of loops formed and the number of pruning events vary from embryo to embryo . To estimate the average number of loops , we analyzed 19 SPIM movies ( each 40–50-h long ) and quantified the number of pruned , closed by collateral fusion , and remaining loops as well as the overall loop number in each SIV plexus ( Fig 1F and S1 and S2 Movie ) . We observed a total of 74 loops , with an average of ~4 ±1 . 5 loops per plexus . Fifty ( ~67% ) of the loops were eventually removed by regression of the cross branch , 14 ( ~20% ) were closed by collateral fusion of the cross branch to a neighboring major branch , and 10 ( ~13% ) remained until the end of the monitoring time . From these results , we conclude that blood vessel regression is the preferred pruning mechanism during plexus remodeling in the SIV . To determine whether blood flow is important for remodeling of the SIV plexus , we analyzed embryos injected at the single cell stage with the tnnt2/silent heart ( sih ) morpholino [19] , which leads to a lack of heart beat and blood flow during later development ( Fig 1A”–1E” ) . Compared to control embryos , sih morphants do not differ significantly in the number of loops formed ( 36 loops in nine movies , with an average of 4 ±2 per SIV plexus ) , indicating that the outgrowth and the sprouting phases do not require blood flow . Nevertheless , the remodeling of the plexus was strongly affected by the lack of flow and only 4 ( 11% ) of cross branches regressed , whereas 13 ( 36% ) were removed by vessel collateral fusion and 19 ( 53% ) remained until the end of the observed period ( compared to 13% in wild-type embryos ) ( Fig 1F and Fig 1G and S1 Table ) . Previous reports suggest that in certain cases pruning is accompanied by endothelial cell apoptosis [12 , 20] . We used a transgenic line labeling cell nuclei to verify whether apoptotic cells are present during the SIV plexus remodeling . In our time-lapse movies ( S6 Movie , n = 10 ) , we did not observe dying cells near the pruning vessels , and we noted an increase in nuclei number in analyzed pruning regions—a result of observed cell divisions ( Fig 2 and S2 Table ) . The nuclei movements suggested that cell rearrangements , rather than apoptosis , account for vessel regression . The nuclei moved away from the pruning segments and incorporated into the remaining major SIV branches ( Fig 2 and S6 Movie ) . We have previously described vessel sprouts anastomosis in different vascular beds and have defined a highly stereotypic multistep process , which can follow one of two modes depending on either the presence or the absence of blood flow in the fusing sprouts [13 , 14] . To investigate whether these two modes of vessel anastomosis also occur during the formation of the SIV , we performed live imaging of early SIV development in a transgenic line , ( Tg ( fliep:GFF ) ubs3 , ( UAS:mRFP ) , ( 5xUAS:cdh5-EGFP ) ubs12 ) , which labels cell–cell junctions by expressing green fluorescent protein ( GFP ) fused to a “tailless” version of vascular endothelial ( VE ) -cadherin . We observed that SIV sprouts follow the fusion steps as previously described . The early sprouts were not lumenized and formed anastomoses in a nonperfused environment . After the flow entered the developing SIV ( ~40 hpf ) , sprouts became lumenized and followed the steps of anastomosis previously observed in a perfused environment , including the formation of seamless tubes and the subsequent transformation into a multicellular tube ( S1B Fig and S1C Fig and S4 Movie ) . Maturation of the SIV requires remodeling of the initial complex network in order to reach its mature topology , consisting of vertical , parallel branches connecting dorsal and ventral longitudinal vessels . In our long-term SPIM-imaging analyses , we found that remodeling is mostly achieved through regression of supernumerary branches formed during the SIV sprouting phase . The significant number of pruning events in each SIV plexus ( 2 . 5 ± 1 . 5 , n = 19 ) and the convenient positioning of the SIV on top of the yolk allowed us to perform single cell analyses in great detail , using transgenic lines labeling cell–cell junctions ( S7 Movie ) . We found that , similarly to vessel fusion , pruning occurs in two variations characterized by the absence or presence of flow in the pruning branches during cell rearrangements , which we named pruning type I and II , respectively . In pruning type I , the first recognizable step was the collapse of lumen in the vessel branch , which was transformed this way into a nonlumenized , multicellular cord with continuous junctional connections , often visible as two parallel lines ( Fig 3A1–3A3 and S8 Movie ) . Subsequently , cells moved away from the pruning branch and incorporated into the neighboring major branches , eventually leaving a single bridging cell in between the latter ( Fig 3A3 and Fig 3A4 ) . This way , the vessel architecture changed from multicellular to unicellular , as visualized by changes in the junctional pattern . Initially , the junctions were continuous ( Fig 3A1–3A3 , green arrow ) , and after cell rearrangements , we observed ajunctional areas indicating unicellular vessel segments ( Fig 3A4 , grey arrow ) . During this transformation , the junctions of the last bridging cell changed from continuous along the cell body into two separate junctional contacts on both poles of the cell . With those junctions , the bridging cell was connected to the opposing major branches ( Fig 3A4 ) . Eventually , one of these connections shrank to a single cytoplasmic extension when the cell body migrated and incorporated into the opposing major branch ( Fig 3A5 and S8 Movie ) . Finally , this last contact was resolved and regression completed ( S3 Fig and S9 Movie ) . In pruning type II , similar cellular rearrangements were observed , but the lumen was maintained in the process almost until the end . Initially , the junctions were continuous in the lumenized tube ( Fig 3B1 and Fig 3B2 , green arrow ) . Cell rearrangements were reflected in the change of the junctional pattern , and ajunctional areas appeared when only a single , bridging cell remained in the regressing branch ( Fig 3B3 and Fig 3B4 , white arrow ) . In this case , since the lumen was maintained , the last bridging endothelial cell transformed its architecture into a unicellular tube ( Fig 3B3 ) . The transcellular lumen of the unicellular tube collapsed , and the continuous apical surface within the cell was separated into two compartments ( Fig 3B4 ) . After the lumen collapsed completely , the cell body moved towards one of the major branches , shrinking its contact surface with the opposite major branch to a single spot ( Fig 3B5 and 3B6 ) . This last contact was eventually resolved as the cell completely incorporated into the opposite major branch ( Fig 3B6 and S10 Movie ) . We analyzed a total of 57 pruning events in transgenic lines marking endothelial cell–cell junctions and cell cytoplasm ( Tg ( fliep:GFF ) ubs3 , ( UAS:mRFP ) , ( 5xUAS:cdh5-EGFP ) ubs12 , n = 29 , S7 Movie ) or cell membrane ( TgBAC ( kdrl:mKate-CAAX ) ubs16 , n = 28 , S11 Movie ) . In the first case , we were able to assess both the cell arrangements and the presence of lumen during blood vessel regression . By labeling endothelial cell membranes with mKate2-CAAX , we were able to follow the dynamics of apical membrane , which is labeled more strongly than basal membrane and highlights inflated luminal compartments in unicellular blood vessels and cell–cell contact surface of multicellular vessels ( S4 Fig and S11 Movie ) [14] . From these experiments , we conclude that at least 30% of the pruning events involved formation of transient unicellular , lumenized tubes ( Fig 3C ) . Our time-lapse analyses revealed that type II pruning involves the formation of transient unicellular tubes formed by a single endothelial cell , which wraps itself around the lumen . Surprisingly , upon contact with its contralateral side the cell starts to self-fuse its cell membrane in a zipper-like fashion , thereby transforming the cell into a seamless , unicellular tube with a transcellular lumen ( Fig 4 ) . To demonstrate endothelial cell self-fusion more directly , we performed single cell labeling experiments with the junctional marker EGFP-ZO-1 ( Tg ( fliep:GFF ) ubs3 , ( UAS:mRFP ) , ( UAS:EGFP-ZO-1 ) ubs5 ) , which tends to be expressed in a mosaic fashion [13] . Since this transgene is hardly active in venous vessels , we examined whether endothelial cell self-fusion occurs during regression of segmental arteries ( SeAs ) , which takes place during segmental vein formation ( SeV ) in the trunk . SeVs sprout from the posterior cardinal vein and form by anastomosis with SeA , thereby transforming the latter into SeVs [21] . This fusion event is accompanied with the regression of the proximal segment of the SeA ( S5 Fig and S12 Movie ) . Using the UAS:EGFP-ZO1 marker , we were able to follow a single endothelial cell as it formed a unicellular tube ( Fig 4 and S13 Movie ) . The junctional transformation of this cell evidenced by ZO-1 expression was entirely consistent with endothelial cell self-fusion . Initially , the SeA was connected to the dorsal aorta with at least two cells , one of them labeled with EGFP-ZO-1 ( Fig 4A , green cell ) . The arterial segment , initially multicellular , undergoes rearrangements similar to the ones described for the SIVs , also in two possible variations ( S5 Fig ) . Fig 4 shows pruning type II , in which lumen was maintained during cell rearrangements and the remaining “last bridging” cell wrapped around the lumen and self-fused to form a unicellular tube . The cell had a funnel-like shape and connected the aorta ( bottom , large junctional ring ) to the SeA ( top , small junctional ring ) . Eventually , the SeA connection to the aorta was resolved , and the cell incorporated completely into the aorta ( S13 Movie ) . In type II regression , the vascular lumen collapses in a unicellular context . Here , the continuous apical compartment lining the luminal site of the cell was separated into two compartments when the opposing apical membranes collapsed on each other because of lumen deflation ( Fig 5 ) . Interestingly , in many cases the lumen split right next to the nucleus , where the cell body takes up the most space ( Fig 5A and Fig 5C1 , asterisks ) , thereby facilitating the separation of luminal compartments ( Fig 5B and S14 Movie ) . Even though in several cases it took up to 12 hours to complete this step of the pruning process , the lumen collapse itself was very fast . When we analyzed luminal membrane using high-resolution time-lapse imaging , we found that the lumen broke and reconnected multiple times before completely separating the two remaining luminal compartments ( Fig 5C and Fig 5D and S15 Movie ) .
Developmental pruning of vascular networks has been described in various vertebrate organs , most notably in the extraembryonic tissues of the avian embryo and the postnatal mouse [22] . Here , we describe the SIV plexus of the zebrafish embryo as a novel in vivo model to study angiogenesis and vascular remodeling . Using various transgenic reporters and cutting-edge microscopy techniques , we first describe vascular remodeling on a plexus-wide scale and then turn to the analysis of single cell behaviors to uncover the cellular mechanisms of blood vessel regression . Vascular remodeling is an extremely dynamic process that is regulated by molecular as well as physical cues [6 , 7 , 23 , 24] . While the molecular regulation has been studied in mouse retina [6 , 7 , 24] , avian chorion-allantoic blood vessels have been used to study the dynamics of vascular remodeling [9] . In zebrafish , the blood vessel regression has previously been studied in the mid-brain [11] , the eye [12] , and during intersegmental vessel formation in the trunk of the embryo [21] . In the latter two contexts , pruning appears to be triggered by a fusion event between an angiogenic sprout and an existing vessel . This fusion event generates a “T-junction , ” which subdivides the existing vessel into two segments of different flow patterns , which ultimately lead to the regression of one of the segments . This is in contrast to the pruning patterns we observe in the SIV . Here , the blood vessels first form a reticular network and then remodel through pruning of supernumerary segments . This temporal separation of angiogenesis and pruning resembles formation of the murine retinal vasculature . In the retina , sprouting and fusion of vessels takes place at the angiogenic front , whereas pruning is observed in proximal regions [7 , 25] . Therefore , the SIV plexus of the zebrafish provides an in vivo model well suited to complement studies of vascular remodeling in the mouse retina , with the added benefit of accessibility for live imaging . In all instances that we observed , pruning involved extensive rearrangements of endothelial cells , which moved out of the branch and incorporated into the neighboring vessels . Furthermore , we did not observe cell death or hemorrhages during this process . In agreement with previous reports [11 , 12] , we observed a reduction and eventually the halt of blood flow in pruning branches , prior to their disassembly . The lumen , however , was often maintained and persisted in endothelial remodeling up to the point when only a single cell connection was remaining . In fact , the choice between the two morphogenetic pruning pathways ( multicellular versus unicellular ) correlated with the timing of lumen collapse relative to the initiation of the cell rearrangements . This suggests that endothelial cells adapt their morphogenetic behavior to the constraints imposed by the state of the vascular lumen at the onset of blood vessel pruning . Based on our observations , the selection of the pruning mode I or II is a consequence of the cellular topology and flow environment of the pruning segment . The primary question is how the segment to be pruned is selected in the first place . Genetic factors such as Notch/Dll4 [6] and Wnt [26] have been shown to influence vessel remodeling in the developing mouse retina . Importantly , reduction in Notch/Dll4 signaling led to an up-regulation of vasodilators such as adrenomedullin and prevented retinal capillary regression [6] , thus suggesting that Notch/Delta signaling may provide the switch for the selection of pruning blood vessels by locally modulating blood flow patterns . The role of Wnt signaling in blood vessel pruning is less clear . While several Wnt ligands and receptors have been implicated in blood vessel regression [26] , a recent study by Korn et al . [24] has investigated the role of noncanonical Wnt signaling by endothelial specific knock-out of Evi , which is required for the secretion of Wnt ligands . Such Evi knock-out retinas not only display strongly enhanced blood vessel regression in the proximal ( postangiogenic ) region of the retina but also affect blood vessel density at the angiogenic front . Furthermore , and in contrast to our observations , these defects could be attributed to a decrease in endothelial cell proliferation and an increase in apoptosis . However , recent studies on blood vessel regression in the retina ( see accompanying paper by Franco et al . [27] ) as well as our observations do not support apoptosis as a key regulator of blood vessel regression . These discrepancies suggest that noncanonical Wnt signaling is not directly involved in blood vessel selection for pruning . The observation that changes in blood flow patterns precede blood vessel regression suggests that mechanical forces are essential regulators of pruning . In agreement with this notion , blood pressure , shear stress , and flow type were shown to affect vessel remodeling [28 , 29] . When the flow was disrupted , vessels did not remodel properly . Interestingly , the remodeling involved detachment of small branches defined by differences in flow and the angle of branch bifurcation [30] . Also in the zebrafish brain vessels , pruning patterns could be altered by changes in blood pressure . Pruning was induced in branches with artificially halted flow and inhibited when blood pressure was chemically raised [11] . In the mouse yolk sac vasculature , erythroblast circulation was required for remodeling and for the expression of force-regulated factors , pointing at the role of blood viscosity [10] . Our results show that in the SIV mostly the small , bifurcated branches are removed and that instable blood flow and lumen are the first signs of pruning , suggesting that a similar flow-dependent selection mechanism could be involved . This view is supported by our observation that pruning in the SIV is impaired in the absence of blood flow , suggesting that certain flow-dependent factors , likely the differences in pressure between branches , trigger pruning . Upon initiation of blood vessel regression , endothelial cells migrate out of the regressing branch . This behavior is regulated by a change in cell polarity , which itself appears to be controlled by hemodynamic forces ( see accompanying article [27] ) . Our analyses are in agreement with this view and show that these cell rearrangements and especially the lumen collapse are highly dynamic processes that are strongly influenced by the presence of blood pressure in the developing vessels . During SIV pruning , endothelial cells can embark on two different morphogenetic pathways , of which only type II pruning involves cell wrapping and cell self-fusion ( Fig 6 ) . At this point , we do not have evidence whether a particular signaling pathway regulates the choice of the morphogenetic pathway . Type I and type II pruning mechanisms split up only after endothelial cells have initiated cell rearrangements , suggesting that neither Notch/Dll nor noncanonical Wnt signaling is directly involved in the determination of type I versus type II specification . In contrast , type I/II pruning strongly correlates with the presence ( type II ) or absence ( type I ) of lumen during this process . This correlation suggests that pruning type selection is rather controlled by mechanical cues and that endothelial cells adapt their behavior to different luminal topologies during cell rearrangement . Apart from blood vessel regression through pruning , we observed an alternative mechanism for reducing the number of vascular loops , namely collateral fusion . Morphologically , collateral fusion resembles reversed blood vessel intussusception , which is an angiogenic process in which vessel plexus expands by splitting existing branches along the longitudinal axis . Intussusception is induced in response to high flow and therewith-increased wall shear stress [28 , 31] . Interestingly , in the SIVs we observed an increase of collateral fusion events in the embryos lacking blood flow and hence pruning . Further studies will show whether these processes are regulated and coordinated by the blood-flow-related factors . In this study , we demonstrate that blood vessel pruning is a stereotyped , stepwise process . On a cellular level , vessel pruning resembles vessel fusion , but many of the cell activities happen in a reverse order . However , the molecular mechanisms employed for certain cellular activities in a reverse order should be quite different . In both fusion and pruning , we observed changes in cell polarity , dynamic formation and resorption of apical/luminal membrane , and dynamic cell rearrangements that impose dramatic shape changes on endothelial cell architecture . Interestingly , the presence or absence of luminal pressure defines the mode of blood vessel fusion as well as pruning . A critical step during blood vessel fusion is the formation of a continuous lumen , which can happen either through cell rearrangements and formation of a multicellular cord ( in a nonperfused environment ) or through pressure-dependent apical membrane invagination and formation of a seamless unicellular tube ( in a perfused environment ) [13 , 14] . Analogously , during vessel pruning , we observe a reversed process in which a continuous luminal compartment within the “last bridging” cell is partitioned into two separate ones . Also in this case , we observed two different scenarios . In type I pruning , analogous to fusion in a nonperfused environment , cell rearrangements following lumen deflation lead to the subdivision of the continuous apical compartment of the last linking cell into two individual compartments . This becomes apparent when continuous junctions are separated into two independent rings on the opposing ends of the cell ( Fig 6B’–6D’ ) . In type II pruning , analogous to fusion in a perfused environment , a seamless tube forms as a result of cell rearrangements; subsequently , the lumen collapses within this unicellular tube , leading to fission of apical membranes and generation of two separate luminal compartments ( Fig 6C”–6 D” and S6 Fig ) . During anastomosis as well as during regression , a blood vessel often goes through a transient unicellular state , which is marked by unsteady and intermittent blood flow . During this phase , the luminal/apical compartment collapses ( forming transient vacuole-like structures ) and reforms in rapid succession . On the one hand , this shows that continuous blood pressure is required to stabilize the luminal compartment within a unicellular tube . On the other hand , the reiterative splitting and rejoining of apical membrane compartments within endothelial cells illustrates an extensive capacity of apical membrane for budding and fusion—presumably mediated by factors implicated in vesicle trafficking and fusion [32] . During the pruning process , we find that the transformation of a multicellular tube into a unicellular tube results from one endothelial cell wrapping around the lumen and establishing a self-contact , which ultimately results in membrane fusion at the lateral side of the cell and formation of a seamless tube . This process of cell self-fusion resembles cell splitting in reversed order , a process we have previously described as occurring during blood vessel fusion , in which an initial unicellular tube transforms into a multicellular tube and the cell shape undergoes a reversed transformation , from a seamless tube cell into a flat cell [14] . Even though geometrically similar and reversed , these two processes must involve very different molecular mechanisms , since different membrane compartments interact with each other in each case ( S6 Fig ) . During cell splitting , two inner leaflets of opposing lateral membranes approach each other and must be the first to fuse . This process could be mediated by intracellular components , for example , similar to those involved in vesicle fusion or those involved in cell abscission—the final step of cytokinesis [33] . In cell self-fusion , a cell wraps itself around a lumen until the two outer membrane leaflets are connecting . Such a cell behavior has been previously described in the tracheal system of Drosophila [34] . Here , cell wrapping results in a unicellular tube sealed by an autocellular adherens junction . During endothelial cell wrapping , an initial autocellular contact is made , but it does not result in an extensive autocellular junction . Instead , the cell–cell contact is followed by membrane fusion . It will be interesting to decipher the molecular mechanisms that allow or prevent cell self-fusion in endothelial and tracheal cells , respectively . Cell fusion is a rather prominent occurrence during development—for instance , it occurs in the fusion of gametes , the generation of multinucleate muscles , and placenta formation [35] . Cell self-fusion is a much more rare phenomenon but has already been described in the formation of tubular structures in Caenorhabditis elegans [36 , 37] . During the formation of the digestive system , several cells autofuse to generate toroid structures , similar to what we observed during vascular pruning . In the case of the C . elegans intestinal tract , autofusion ( self-fusion ) is mediated by so-called fusogens ( AFF-1 and EFF-1 , respectively ) , which help to bring adjacent membranes in close proximity to initiate membrane fusion [36] . It remains to be investigated whether proteins with similar function exist in vertebrates . Cell self-contact elimination ( similar or identical to cell self-fusion ) has recently been reported in epithelial Madin-Darby canine kidney ( MDCK ) cells plated on a micropillar array [38] . In this assay system , cells were forced to grow around the pillars to take up a torus shape , with the pillar in the middle , and to make contacts with themselves at the distal side of the pillar . Such cell self-contacts resulted in cell self-fusion , promoted by the presence of E-cadherin , which brings cell membranes into proximity and thus facilitates membrane fusion . Strikingly , cells lacking E-cadherins , such as fibroblasts , failed to self-fuse . This suggests that cell self-fusion is an inherent feature of epithelial cells that have to ensure continuity of the tubular organ lining . To the best of our knowledge , our study is the first report demonstrating that cell self-fusion is an integral part of the complex morphogenetic process during vascular pruning in vertebrates .
The animal experiments were approved by the Kantonales Veterinaeramt Basel-Stadt ( license number 1995/1996 ) and in accordance with EU directive 2011/63/EU as well as the German Animal Welfare Act . Zebrafish were maintained at standard conditions [39] . Embryos were staged by hpf at 28 . 5°C [40] . The following zebrafish lines were used in this study: Tg ( fli1a:EGFP ) y1 [41] , Tg ( kdrl:EGFP ) S843 [42] , Tg ( kdrl:EGFPnls ) UBS1 [43] , Tg ( UAS:EGFP-ZO1-cmlc:EGFP ) UBS5-7 [13] , Tg ( UAS:RFP ) and Tg ( fli1ep:GAL4FF ) UBS2-4 [44 , 45] , Tg ( 5xUAS:cdh5-EGFP ) ubs12 [14] , Tg ( BAC:kdrl:mKate2-CAAX ) UBS16 [14] , and Tg ( UAS:Kaede ) rk7 [46] . Tnnt2/silent heart ( sih ) Morpholino injection was performed as described before [19] . Transgenic embryos selected for presence of fluorescence were anaesthetized in 1x tricaine ( 0 . 08% ) and mounted in a 35-mm glass-bottom petri dish ( 0 . 17 mm , MatTek ) , using 0 . 7% low-melting agarose ( Sigma ) containing 0 . 08% tricaine and 0 . 003% PTU . A Leica TCS SP5 confocal microscope was used for time-lapse analyses . Images were taken using the following objectives: 20x air or 40x water immersion . High time-resolution analyses were performed using a Perkin Elmer Ultraview spinning disc microscope and a 63x water immersion objective , and images were deconvolved using Huygens Remote Manager software [47] . All images are maximum intensity z-projections . Photoconversion of the Kaede protein was performed as described before [13] . The posterior cardinal vein was selected as a region of interest and illuminated with a 405-nm solid-state laser for ~30–40 s to achieve different color labeling of the venous ( red ) and arterial ( green ) cells . To avoid pigmentation , transgenic zebrafish embryos were kept in 0 . 2-mM N-Phenylthiourea ( PTU ) from 14 hpf until 35 hpf . At 35 hpf , transgenic zebrafish embryos were selected for the presence of fli1a:EGFP fluorescence and mounted in fluorinated propylene ethylene ( FEP ) tubes , which had been coated with 3% methyl cellulose and filled with 0 . 1% low-melting agarose [48] . To prevent movement of the fish , the chamber of the SPIM setup was filled with E3 fish medium containing 0 . 03% tricaine . Long-term time-lapse imaging was performed on a home-built multidirectional SPIM ( mSPIM ) setup [49] , equipped with a Leica HCX APO L 20x/0 . 50 W detection objective , a Coherent Sapphire 488-nm laser , and two Andor iXon 885 . Both SIV plexuses of the embryo were imaged every 15 min for up to 60 h ( S1 Table and S1 Data ) . The images were acquired from one angle for each SIV with an axial resolution ( z-stack spacing ) of 3 μm . To image the whole plexus length , several regions were acquired and stitched together using an adapted version of the stitching tool [50] . The acquired images were processed and background was subtracted using available and custom written Fiji plugins [51] . The side projections were generated using thresholding by Huang [52] and subsequent Chamfer Matching [53] . ISVs were used as a reference point to follow the embryo growth and adjust the side-profile absolute position over the time-lapse to achieve an optimal projection . To generate the 3-D rendering in S2 Movie , the 3-D project plugin ( routine written by Michael Castle and Janice Keller of the University of Michigan Mental Health Research Institute [MHRI] ) of ImageJ was applied using brightest point projection , interpolation , and angle increment of 10 degrees . The quantification of pruning events in long-term SPIM movies was performed on maximal projections . Vascular structures generated between one or two major branches and a small branch were counted as vascular loops . The number of pruned , closed by collateral fusion , and remaining loops was counted for each experiment throughout the time lapse ( S1 Table ) . Average values were compared between wild type and silent heart experiments using unpaired , two-tailed Student’s t test ( S1 Data ) . The quantification of nuclei number in time-lapse movies was performed on maximal projections , within a region of interest defined around the pruning branches . Nuclei numbers were assessed every 10 time points before , during , and after pruning ( S2 Table ) . Movies were analyzed qualitatively for the presence of apoptotic and dividing nuclei . Nuclei tracking in S6 Movie was performed using Imaris Spot Tracking tool ( Bitplane ) . | The blood vasculature circulates gas , nutrients , hormones , and metabolites to all organs of the body . It is indispensable for survival and already functions at very early stages of embryonic development . At this point , new blood vessels form mainly through angiogenesis—the outgrowth of new vessels from existing ones . New vascular sprouts connect to each other to form functional loops with blood flow , a process termed anastomosis . Vascular plexuses formed in this way subsequently remodel to a final structure with efficient flow patterns . Remodeling often involves pruning or regression of unnecessary branches , leading to a simplification of the network . Our in vivo live imaging studies of the pruning process in the zebrafish embryo show that vessel regression occurs through cell rearrangements , wherein cells consecutively migrate out of the pruning branch . As a result , the initially multicellular vessel is reduced to a single cell connection that is eventually resolved when the last cell incorporates into the neighboring branch . If the lumen is maintained in the pruning vessel , the process involves transient formation of a unicellular tube through cell self-fusion . Thus , we show how a variety of cellular activities are coordinated to achieve vessel pruning . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| []
| 2015 | Endothelial Cell Self-fusion during Vascular Pruning |
The pattern and frequency of insertions that enable transposable elements to remain active in a population are poorly understood . The retrotransposable element R2 exclusively inserts into the 28S rRNA genes where it establishes long-term , stable relationships with its animal hosts . Previous studies with laboratory stocks of Drosophila simulans have suggested that control over R2 retrotransposition resides within the rDNA loci . In this report , we sampled 180 rDNA loci of animals collected from two natural populations of D . simulans . The two populations were found to have similar patterns of R2 activity . About half of the rDNA loci supported no or very low levels of R2 transcripts with no evidence of R2 retrotransposition . The remaining half of the rDNA loci had levels of R2 transcripts that varied in a continuous manner over almost a 100-fold range and did support new retrotransposition events . Structural analysis of the rDNA loci in 18 lines that spanned the range of R2 transcript levels in these populations revealed that R2 number and rDNA locus size varied 2-fold; however , R2 activity was not readily correlated with either of these parameters . Instead R2 activity was best correlated with the distribution of elements within the rDNA locus . Loci with no activity had larger contiguous blocks of rDNA units free of R2-insertions . These data suggest a model in which frequent recombination within the rDNA locus continually redistributes R2-inserted units resulting in changing levels of R2 activity within individual loci and persistent R2 activity within the population .
The abundant transposable elements present in eukaryotes have over the course of evolution helped shape the size , structure and expression of their genomes . Control over the spread of transposable elements has been suggested to be a consequence of the harmful effects of their insertions [1] , [2] and involves either the epigenetic regulation of heterochromatin formation to block transcription [3] or small RNA pathways to degrade their RNA transcripts [4] . In spite of these controls , in most populations new insertions occur continuously at low levels . In Drosophila melanogaster for example , nearly half of spontaneous mutations are caused by transposable element insertions [5] , and the insertion locations of most transposable elements are polymorphic in a population [1] . Multiple attempts have been made to monitor the activity of specific transposable elements in D . melanogaster and D . simulans [6]–[11] . However , it remains uncertain as to whether transposable elements maintain themselves by the inability of the control mechanisms to prevent all events , or the loss of transposable element control in a fraction of the individuals in a population . The ribosomal RNA ( rRNA ) genes of eukaryotes are composed of hundreds to thousands of tandemly repeated units ( the rDNA loci ) . Mature 18S , 5 . 8S and 28S rRNAs are processed from the single transcript of each rDNA unit . Frequent recombination ( unequal crossovers ) within these rDNA loci removes sequence variation , a process referred to as concerted evolution [12] . Concerted evolution is so efficient almost no nucleotide sequence variation exists between the different rDNA units of a locus [13] , [14] . Given this ability of the locus to rid itself of variation , it is surprising that the rDNA locus is home to many specialized transposable elements [15]–[17] . For example , R1 and R2 are non-LTR retrotransposable elements that insert specifically into the 28S rRNA genes ( Figure 1 ) . R1 and R2 are highly adept at maintaining themselves in the rDNA locus . They are found in most arthropods and appear to have been vertically inherited since the origin of the phylum [18] , [19] . Remarkably , R2 elements appear to have been inserting into the same site of the large subunit rRNA gene since near the origin of metazoans [17] . Within many insect species a large fraction of the rDNA units are inserted by R1 and R2 suggesting rapid rates of insertion [20]–[22] . Even though these many inserted units can not make functional 28S rRNA , the host does not appear unduly affected because in most organisms many more rDNA units are encoded than are needed for the production of rRNA [23] . An analysis of inbred laboratory stocks of D . simulans originally derived from a population in Paradise , CA revealed lines with no R2 activity as well as lines with extremely high rates of R2 retrotransposition [24] , [25] . The R2 retrotransposition activity was found to be under transcriptional control with the genetic differences between active and inactive stocks closely linked to the rDNA locus [26] . Minor , if any , influence on the level of R2 activity was found associated with the autosomes . Analysis of the rDNA loci from the active and inactive stocks revealed that the numbers of R2 elements in the active stocks were on average twice that found in the inactive stocks . The R2 elements in the active stocks were distributed throughout the rDNA locus suggesting a model in which a high density of R2 elements prevents the host from activating the needed number of uninserted rDNA units without also activating R2-inserted units . Because the active and inactive stocks used in these studies had been maintained in the laboratory for over 10 years , and considerable changes in the structure of the rDNA locus had occurred during that time [25] , it was unclear whether R2 activity in these laboratory stocks were an accurate reflection of the levels and pattern of R2 activity in natural populations . Here we studied R2 element activity in two populations of D . simulans immediately after their capture . A 100-fold range of R2 transcript levels was detected and correlated with R2 retrotransposition activity . However , the number of R2 elements in the rDNA loci of the active and inactive lines did not substantially differ , suggesting the previously detected accumulation of R2 in active stocks was the result rather than the cause of R2 activity . The property of the rDNA locus that best correlated with R2 activity was subtle changes in the distribution of elements within the rDNA locus .
R2 transcript levels were monitored in RNA isolated from adult females the first generation after establishing the iso-rDNA lines . To enable comparisons between different blots , RNA from line 58 , a laboratory stock previously shown to have stable high levels of R2 transcription and retrotransposition , was included in each analysis [25] , [26] . A representative RNA blot is shown in Figure 2 . Panel A is the RNA probed with R2 sequences , panel B is the same blot probed with a control gene ( alcohol dehydrogenase ) , and panel C is the ethidium bromide staining pattern of the RNAs used . The 3 , 600 nt hybridizing band detected at various intensities in panel A represents full-length R2 element transcripts [26] . Because the hybridization probe used in the blot was from the 5′ end of the R2 element ( Figure 1 , probe 1 ) , the series of lower hybridizing bands represent intermediate degradation products of full-length transcripts , rather than the transcripts from 5′ truncated R2 elements . These lower hybridizing bands between lines generally corresponded in relative intensity to the 3 , 600 nt band suggesting the lines had similar rates of R2 RNA degradation . Shown in Figure 3 is a summary of the R2 transcript levels in all 180 iso-rDNA lines . The populations from both Atlanta and San Diego had transcript levels that varied over a wide range . Both populations also had similar fractions of the lines at comparable transcript levels . About 45% of the stocks derived from the San Diego population and 60% of the stocks from the Atlanta population had no or extremely low levels of R2 transcripts ( defined here as hybridization less than 5 times the background hybridization to the filters ) . The remaining lines from each population had R2 transcript levels that varied in a continuous manner from 6 to 85 times background . Only four lines ( two from each population ) had transcript levels similar to or greater than the R2 active laboratory stock , line 58 . To determine if the R2 transcript levels rapidly changed while being maintained in the laboratory , 32 lines that spanned the full range of R2 transcript levels from the two populations were re-assayed either 2 or 3 generations after the initial screening . No line changed significantly in transcript levels . Lines with intermediate levels of transcription showed the highest levels of change between the two generations with some lines increasing and others decreasing by 20% ( data not shown ) . Nine lines from each population were selected to represent the observed range of R2 transcript levels ( lines indicated with asterisks in Figure 3 ) . Because the flies for each population were collected from one location over a period of only a few days , some of the flies could be closely related and thus not represent independent rDNA loci . To determine the relatedness of the rDNA loci from these lines , the specific R2 insertions in each of the 18 lines were compared . R2 elements generate identical 3′ junctions with the 28S gene but about one-half of the copies have variable 5′ junctions [27] . This 5′ end variation predominantly corresponds to truncations and has been suggested to arise from the R2 polymerase failing to reach the 5′ end of the R2 transcript during reverse transcription and/or from DNA repair processes associated with partial retrotransposition events [15] . The 5′ ends of the R2 elements from each line were amplified in a series of PCR reactions , and the products separated at single nucleotide resolution on high-resolution sequencing gels . These R2 5′ truncation profiles serve as a robust means to determine whether the rDNA loci of the different lines are related . For example , laboratory stocks with inactive R2 elements continue to have identical 5′ truncations after 100 generations [24] , [27] , while 60–90% of the truncations are identical after 30 generations when the R2 elements are active [25] . Shown in Figure 4 are the collections of 5′ truncated R2 insertions found in the 18 lines . The distances from the vertical lines to the 3′ end of the element represent the lengths of the different 5′ truncated elements in each line . Each line had from 13 to 28 ( mean 20 ) different length 5′ truncated copies . From 2 to 9 ( mean 6 ) of these elements were shared with one or more other lines ( identified in the figure with dotted vertical lines ) . These shared copies consisted largely of six specific truncations ( vertical dotted lines marked at the top with asterisks ) that were also shared between the Atlanta and San Diego populations . These long-term stable R2 copies are presumably located in regions of the rDNA locus , perhaps the edges , which seldom undergo recombination events . These data indicate that the rDNA loci in the 18 lines selected for further study are not closely related with their rDNA loci having undergone significant turnover of R2 elements since their common ancestor . Finally , the 5′ truncation profile of each line also differed from the rDNA loci present in the Bx line ( not shown ) , indicating no instance of recombination between the Bx marker and the rDNA locus during the establishment of the iso-rDNA lines . R2 activity ( retrotranspositions ) was next scored in the 18 representative lines . To monitor new retrotransposition events , the R2 5′ truncation profiles of 16 males from each line in the eighth generation after formation of the iso-rDNA lines were compared to the profile of the original male used to establish each line . Retrotransposition ( insertion ) events were scored as additional R2 5′ truncations observed in any of the male progeny . Because R2 deletions have also been correlated with R2 retrotransposition activity [25] , R2 copies present in the original rDNA loci but missing in the eighth generation loci were also scored as R2 activity . R2-induced deletions typically involve multiple rDNA units [25] , thus all deletions within a loci were scored as single events , independent of the number of R2 copies involved . Previous analyses have suggested that the insertion and deletion of 5′ truncated copies of R2 are similar to full-length copies , thus these 5′ truncated events represent about half of the total retrotransposition events within the locus [10] , [24] . A summary of the retrotransposition activities of the 18 lines plotted versus their relative R2 transcript levels is shown in Figure 5 . No changes in the 5′ truncation profiles were scored in the six low transcript lines . The 12 lines with readily detected levels of R2 transcripts had on average 4 events ( range 0 to 22 ) . For comparison our laboratory stocks with the highest levels of R2 retrotransposition typically generated 20–30 events in similar experiments [24] , [25] . While not all lines with high levels of transcript had measurable levels of retrotransposition , R2 transcript levels were positively correlated with R2 retrotransposition ( Spearman rank correlation test , r = 0 . 71 , p = 0 . 0005 ) . If the one line with the highest level of activity ( CS24 ) was excluded from the analysis , there still remained a positive correlation ( r = 0 . 67 , p = 0 . 001 ) . Re-examination of the high line ( A141 ) in Figure 5 with no retrotranspositions revealed one retrotransposition event . The lower levels of retrotransposition in some lines with significant R2 transcript levels suggest there may be other levels of control over R2 activity besides transcription . Because of the large number of R2 events that had accumulated in the most active line ( CS24 ) , it was possible to score retrotransposition events in this line at earlier generations . Six events were detected in the first generation after founding the iso-rDNA line , and nine events had accumulated by the third generation . Thus , as in our previous studies of inbred laboratory stocks , high retrotransposition activity can be stable over multiple generations . The total number of R2 copies in each of the 18 lines was determined by directly counting their different 5′ ends . In addition to the 5′-truncated copies shown in Figure 4 , about one-half of the R2 elements are full-length ( i . e . extend to the first nucleotide of the consensus sequence ) . Many of these full-length copies can also be individually counted in PCR assays because they too exhibit length variations associated with small deletions of the 28S gene and non-templated nucleotide additions [28] . To estimate the copy number of those elements with identical length 5′ ends , the intensity of each PCR band was calculated with a regression analysis using single copy variants as reference markers [29] . The total number of R2 copies in the 18 lines varied from a low of 33 to a high of 69 , while the number of full-length R2 elements varied from 18 to 34 ( Table 1 ) . Figure 6A plots the number of full-length and thus potentially active R2 copies versus the level of R2 transcription in each of the 18 lines . The six lines with low R2 transcript levels had on average 23 . 3 full-length copies , while the six lines with highest transcript levels had on average 26 . 8 copies . However , a Spearman rank correlation test suggested no correlation between R2 number and transcript level ( r = 0 . 28 , p = 0 . 19 ) . We next determined if R2 transcript levels correlated with the fraction of the rDNA units inserted with R2 . Because the 3′ ends of all R1 and R2 elements are identical , blots of genomic DNA digested with select restriction enzymes place the uninserted rDNA units , the 3′ ends of R1-inserted units and the 3′ ends of R2-inserted units on different sized fragments . These blots when hybridized with a segment of the 28S rRNA gene immediately downstream of the R1 insertion site reveal the percentage of the total rDNA units that are of each class [30] , [31] . To score those R2 copies that are inserted in the same rDNA unit as an R1 element , a second genomic digest was conducted which placed the 3′ ends of R2 elements in single and double inserted units on different sized restriction fragments and probed with the 3′ end of the R2 element ( probe 3 , Figure 1 ) [10] , [24] . The fraction of the rDNA locus inserted with R2 elements ( both full-length and 5′ truncated ) varied from 16% to 28% ( Table 1 ) . Plotted in Figure 6B is the fraction of the rDNA units inserted with R2 versus the R2 transcript levels . About 17% of the rDNA units in the lowest R2 transcript lines were inserted with R2 , while about 23% of the rDNA units were inserted in the highest transcript lines . While there is a positive correlation between transcript level and fraction inserted when all lines are included ( r = 0 . 61 , p = 0 . 003 ) , there was no correlation when the six low transcript lines were removed ( r = −0 . 17 , p = 0 . 30 ) . The total number of rDNA units ( locus size ) in each of the 18 lines were estimated by dividing the number of R2 elements by the fraction of the rDNA units inserted with R2 ( Figure 7 ) . The rDNA loci varied in size from 185 units to 395 units . Shown in Figure 6C is a plot of rDNA locus size versus the R2 transcript levels . While two of the lines with the lowest levels of R2 transcript did have the two largest rDNA loci , there was no significant correlation between rDNA locus size and R2 transcript level across all 18 lines ( r = −0 . 02 , p = 0 . 48 ) . These findings suggest that unlike our study of laboratory stocks , there was no increase in the number of R2 elements in those rDNA lines with high levels of R2 transcripts . The only significant correlation was a small decrease in the fraction of inserted rDNA units in those lines with essentially no R2 transcripts . Finally , it should be noted that R1-inserted units were at low levels in all stocks ( Table 1 ) and their estimated numbers or fraction of the total rDNA units inserted did not correlate with R2 transcript levels ( data not shown ) . Because neither the size nor the composition of the rDNA locus exhibited a dramatic correlation with the level of R2 transcripts , the distribution of R2 insertions within the locus was next determined . The restriction enzyme NotI cleaves near the middle of the R2 element ( Figure 1 ) , but does not cleave R1 elements or the transcribed and intergenic spacer regions of the D . simulans rDNA units . Thus the size of genomic NotI fragments detected in a line reveals the R2-to-R2 spacing within its rDNA locus; however , 3–12 highly 5′ truncated R2 copies in each line ( see Figure 4 ) will not be mapped by this approach . To determine the distribution of R2 elements in the rDNA loci of the low and high transcript lines , nuclei were isolated from 0–23 hr embryos , embedded in agarose , gently lysed and the DNA digested with NotI [26] . The DNA fragments were separated by pulsed-field electrophoresis , transferred to nitrocellulose and hybridized with a DNA fragment from the 18S rRNA gene . The six lines with the highest levels of R2 transcription as well as six lines with the lowest levels of R2 transcripts were used for this study ( Figure 8 ) . The pulse times for electrophoresis were established to optimize the separation of the largest NotI fragments generated from the rDNA loci . Under these conditions NotI fragments below 100 kb are not well separated from the buffer front moving through the gel . Thus those portions of the rDNA locus containing R2 elements separated by less than eight rDNA units ( the average rDNA unit size is about 12 kb ) are not present in the figure . These closely spaced R2-inserted units represent a small fraction of the rDNA locus ( ∼25% ) but a large fraction of the R2 elements ( ∼75% ) . The D . simulans lines with highest levels of R2 transcripts had on average smaller NotI fragments than the lines with low levels of R2 transcripts . This difference in size was significant , whether one used only the largest NotI fragment ( ∼800 kb for the low lines and 540 kb for the high lines , Mann-Whitney U test , p = 0 . 003 ) , or the combined length of the three largest fragments in each line ( ∼1850 kb for the low lines and 1200 kb for the high lines , p = 0 . 004 ) . This finding suggests that of the various properties of the rDNA locus that were measured , the distribution of R2 elements within the locus appeared to best serve as an indicator of whether R2-inserted units were transcribed . However , even this difference in the spacing of R2-inserted units is subtle in that changes in the location of only a small number of R2 elements within the locus could shift the profile from the low to high transcript patterns .
The goal of this study was to apply what had been learned about R2 activity from our studies of laboratory stocks to characterize the pattern of R2 activity in natural populations of D . simulans . In our previous analyses nuclear run-on transcription experiments indicated that control over R2 element activity was at the level of transcription , while crosses between active and inactive lines revealed that this transcriptional control mapped to the site of all R2 insertions , the rDNA locus on the X chromosome [26] . Therefore , in this study iso-rDNA locus lines were established from two natural populations and their levels of R2 transcripts determined within a few generations of isolation . In both populations , about one-half of the lines had levels of R2 transcripts that were very low to undetectable . In the remaining lines , the level of R2 transcript varied in a continuous manner over a wide range . In the 18 lines studied as representative of the R2 transcript range , a positive correlation was detected between R2 transcript level and R2 retrotransposition activity . While this correlation was not absolute , transcript level appeared to serve as a reliable indicator of potential R2 activity and thus a major means of control over R2 activity in a population . Structural analysis of the rDNA loci of the 18 lines indicated that the increased R2 transcription was not a result of the greater accumulation of R2 copies or changes in the size of the rDNA locus . The property of the rDNA locus that best correlated with the level of R2 transcripts was the size of the largest continuous stretch of units free of R2 insertions . This suggestion that it is a difference in the distribution of R2 elements within the rDNA locus that controls R2 transcription adds considerable support to our previously proposed model of R2 regulation [26] . This model was based on several known properties concerning the transcription of rDNA units and their R2 insertions [26] . First , R2 elements do not appear to encode their own promoter; rather their RNA transcripts are processed from a co-transcript with the 28S rRNA gene [26] , [32]–[35] . Second , the size of the rDNA locus is sufficiently large that only a fraction of the rDNA units encoded by an organism are transcribed at any one time [23] , [36]–[39] . Third , electron microscopic observations suggest that transcriptionally active rDNA units occur in contiguous blocks rather than as dispersed units throughout the locus [40] , [41] . Based on these observations we proposed that arthropods have evolved the ability to identify and actively transcribe one or more regions of the rDNA locus that contain the lowest frequency of R2-inserted units [26] . Thus significant transcription of R2 elements occurs when an organism is unable to identify large continuous regions of the rDNA locus free of insertions . In the laboratory stocks used in our previous study , there was on average a two-fold greater number of R2 elements in active lines compared to the inactive lines . As a consequence it was not possible to exclude the model that the greater number of R2 insertions in the rDNA loci led to higher transcription levels . In our natural lines , on the other hand , the number of R2 insertions was similar between the active and inactive lines , suggesting that it is R2 distribution , not number , that determines whether R2-inserted units are transcribed . Our study also suggests that it may be the location of only a few R2 elements that determines the level of R2 activity within a line . A repositioning of only a small number of R2 elements could change the NotI digestion pattern of the low transcript lines to that of the high transcript lines ( Figure 8 ) . In addition , our determination of transcript level was based on the level of full-length R2 transcripts , while the R2 distribution pattern was based on the presence of a NotI site near the middle of the R2 element . Thus the positioning of 5′ truncated R2 copies , that are of a length to retain the NotI site but not able to generate full-length transcripts , within the regions of lowest R2 abundance could explain why some lines low transcript lines ( e . g . CS22 ) have NotI profiles more similar to a high transcript line . Our analysis of the individual R2 elements present in the rDNA loci of each line ( Figure 4 ) revealed that each rDNA locus had different collections of R2 insertions suggesting that R2 elements are continually being gained and lost within a population . The rDNA loci should , therefore , be viewed as a continuously changing landscape with the individual loci shifting between low and high levels of R2 transcription . These continuously changing rDNA loci can explain another difference between natural and laboratory stocks . Of the 15 lines isolated from Paradise , CA and maintained in the laboratory for about 10 years , about one-fourth ( 4 lines ) had extremely high levels of R2 transcripts and retrotransposition [24] , [26] . However , of the 180 lines established from the Atlanta and San Diego populations , only 2% had transcript levels as high as that seen in the laboratory stocks , and only one of these lines had levels of retrotransposition as high as the laboratory stocks . While it is formerly possible that the Paradise population simply had higher levels of R2 activity , it appears more likely that during the prolonged period the Paradise stocks were maintained in the laboratory the inbreeding and relaxed selection conditions permitted the accumulation of the higher levels of R2 elements observed in these active laboratory stocks . Such shifts to high R2 element activity may also occur in the natural populations from Atlanta and San Diego , but such flies would be lost from the populations by natural selection . We have found that laboratory stocks with high levels of R2 activity are readily out-competed when crossed to laboratory stocks containing low levels of R2 activity ( D . Eickbush , unpublished data ) . One important feature of R2 control detected in our previous study of laboratory stocks was the dominance of expression of rDNA loci with low levels of R2 transcripts over loci with high levels of R2 transcripts [26] . In females containing one rDNA locus from a stock with high levels of R2 transcription ( R2 active locus ) and one locus from a stock with no R2 transcription ( R2 inactive locus ) , rDNA units in the R2 inactive loci were preferentially transcribed . We termed this ability to turn off an entire rDNA locus nucleolar dominance , as it appeared similar to the frequent dominance of one rDNA locus in interspecies hybrids [42] . To determine if such dominance could also be detected between our natural rDNA loci , we conducted reciprocal crosses between the most active line in this study , CS24 , to several inactive lines . In the heterozygous female offspring of both crosses the high level of R2 transcript seen from the CS24 loci was significantly turned off , consistent with the dominance of R2 inactive loci in natural populations ( data not shown ) . However , such crosses involving lines with the more typical lower levels of transcripts were often not recessive to R2 inactive stocks . Thus , while more experiments are needed , nucleolar dominance may only influence loci with extremely high levels of transcripts . Based on our survey of the two D . simulans populations we can estimate an R2 retrotransposition rate for the entire population . Comparing the rate of R2 retrotransposition with the level of transcripts ( Figure 5 ) , about 50% of the rDNA loci in the two populations are likely to undergo no retrotransposition . The remaining rDNA loci have a high probability of supporting some level of retrotransposition . For the 12 rDNA loci we sampled from this group , 46 retrotranspositions were detected when 16 chromosomes were monitored in the eighth generation giving a rate of 0 . 034 events/chromosome/generation [46 events / ( 12 lines ×7 generations ×16 chromosomes ) ] . As a lower limit , if we exclude the one extremely active line from this analysis the R2 retrotransposition rate drops to 0 . 019 events/chromosome/generation . Because only 5′ truncated R2 elements , which represent about half the total number of elements , were monitored in our assays , these rates should be multiplied by a factor of two ( range of 0 . 038–0 . 068 ) . Finally , correcting for the fraction of the rDNA loci that have detectable levels of transcription ( 50% ) the total R2 retrotransposition rate for the population can be estimated at 0 . 019–0 . 034 events/chromosome/generation . While this range is only a rough estimate , it is consistent with computer simulations conducted to mimic the types of recombination and rates of retrotransposition likely to be present in the rDNA loci of Drosophila [23] . The findings in this report address one of the major questions concerning the ability of transposable elements to maintain themselves within a species lineage . Does the control mechanism of the cell occasionally breakdown , or is this mechanism simply incapable of preventing all new insertions ? In the case of R2 our findings support the latter possibility . In most organisms , large numbers of R2 elements can be effectively prevented from transcription by expressing the rDNA units in only a small domain that does not contain R2 insertions . However , recombination events ( in particular unequal crossovers ) will expand , contract and rearrange the rDNA units in the locus meaning these domains free of R2 insertion are not permanently stable . Thus it is not a difference in the cellular control mechanism between animals in a population , but rather differences in their rDNA loci that determines whether R2 elements are being transcribed . The long-term stability of R2 elements is assured because even though all copies can be effectively silenced for long periods of time , the transcription of some copies will eventually be resurrected by recombination . Of course transcriptional control may not be the only level at which the activity of R2 elements are regulated . While retrotranspositions have never been detected in a stock with no detectable R2 transcripts , there may not be a complete correlation between the level of R2 transcripts and the rate of retrotransposition . Retrotransposition events were difficult to detect in one of the lines with the highest level of R2 transcripts ( Figure 5 ) . In one of our laboratory stocks , retrotransposition events were more frequent than expected based on R2 transcript levels [26] . Thus additional post-transcriptional steps appear to influence the rate of R2 retrotransposition in D . simulans . Further studies are needed to determine if these steps represents significant control mechanisms over R2 activity . Finally , this population study not only provides additional support for a domain model of transcription in the rDNA locus but also addresses two evolutionary questions raised by our earlier studies . First , why do the different copies of R2 generated by retrotransposition usually remain single-copy ( i . e . are not duplicated by recombination ) until they are eventually lost from the genome [24] , [27] , [35] ? Second , why in long-term studies of specific rDNA loci did the number of uninserted rDNA units change most rapidly than the inserted units [29] ? Both of these finding can be explained if recombination events in the rDNA loci predominantly occur in the insertion free region of the rDNA locus activated for transcription . Thus an understanding of R2 location and control is needed to appreciate the forces at work in the expression and evolution of the entire rDNA locus .
Isofemale stocks of D . simulans collected in San Diego , CA were generated by Peter Andolfatto . D . simulans females of Atlanta population were collected by Todd Schlenke in Atlanta , GA . The Beadex ( Bx ) marker line was obtained from Allen Orr . The single rDNA locus of D . simulans is located on the X chromosome . The rDNA locus of the isolated stocks was followed using Bx , a dominant wing phenotype marker that is the closest available marker to the rDNA locus in D . simulans . First ( Atlanta ) or second generation ( San Diego ) male progeny of collected females were crossed to Bx females . F1 females from these crosses were then backcrossed to the original male . F2 females homozygous for the wild rDNA loci ( XwXw ) were crossed to XwY F2 males to generate the individual lines . Because heterozygous Bx females ( XBxXw ) do not always express the Bx wing phenotype , three single pair crosses between an F2 female and an F2 male were performed . Only if all F3 males from a pair showed the normal wing phenotype was that line considered to contain only the wild-type rDNA locus ( iso-rDNA lines ) . One iso-rDNA locus line was derived from each female collected from the natural population . Because recombination at a low level can occur between Bx and the rDNA locus , the origin of the rDNA loci was confirmed in those lines used for more extensive studies by comparing the R2 5′ truncation profiles of multiple males with the R2 5′ truncation profiles of the original male and the Bx stock . Genomic DNA was extracted from single flies as described by Gloor et al . [43] . To compare the R2 elements between lines as well as count the total number of R2 elements in each line , the 5′ ends of all R2 insertions were analyzed by PCR amplification . The forward primer , 5′-TGCCCAGTGCTCTGAATGTC-3′ , which annealed to 28S gene sequences 80 bp upstream of the R2 insertion site , was 32P-5′-end-labeled and used with the following series of reverse primers that annealed to R2 sequences at various distances from the 5′ end of a full-length element: 3 . 6 kb ( 5′-GTATGGAAATCTATCGAAAGATACT-3′ ) , 3 . 1 kb ( 5′-GTCACCTGCGGCTTCGAATC-3′ ) , 2 . 8 kb ( 5′- CCCCTTGTAGTACGAGACTTC-3′ ) , 2 . 6 kb ( 5′-GCCGGACGCGATAACAATTC-3′ ) , 2 . 0 kb ( 5′-GATAGAAAATCCAACGTTCTGTC-3′ ) , 1 . 6 kb ( 5′-TCGAATGCCTTGCTTACATC-3′ ) , 1 . 3 kb ( 5′-GAAGACGGTTCTGGCCAGTC -3′ ) , 1 . 0 kb ( 5′-CGCTGGACGACAGCATACTGC-3′ ) , 0 . 4 kb ( 5′-CATCAAGTTCGTCTGGGTGC-3′ ) and 0 . 1 kb ( 5′-GACTTGAGTAAAGGAGAGACT-3′ ) . The labeled PCR products were separated on an 8% high voltage denaturing polyacrylamide gel , exposed to a PhorsphorImager screen , and quantitated with a Molecular Dynamics PhorsphorImager scanner using ImageQuant software . Most PCR bands ( est . 80% ) were of equal intensity and corresponded to single copies of the R2 element . To quantify the bands derived from multiple copies of R2 with identical length 5′ ends , the intensity of each PCR band was calculated with a regression analysis using the many single copy variants as reference [29] . While labor intensive , we have found this approach to be more accurate than quantitative PCR or relative hybridizations [24] , [27] , [29] , [35] . To score R2 retrotransposition and deletion events , the upstream ( forward ) primer without end-labeling was used with the above set of reverse PCR primers , separated on 8% native polyacrylamide gels and stained with ethidium bromide . Total RNA was extracted as previously described [35] from 30 adult females of each line in the third generation . Ten micrograms of RNA were separated on 1% agarose , 2 . 2 M formaldehyde gels , the RNA transferred to GeneScreen Plus , and hybridized with an anti-sense RNA probe from the 5′ end of the R2 element ( probe 1 , Figure 1 ) as previously described [26] . The relative levels of 3 , 600 nt full-length R2 transcripts were quantitated on a PhorsphorImager . As a control for RNA loading and quality , all R2 hybridization signals were standardized by monitoring the level of alcohol dehydrogenase hybridization on the same blots [26] . As a control for hybridization efficiency , an equal aliquot of RNA isolated from the previously characterized lab stock , line 58 [24] was included in each blot . Genomic DNA was extracted from 30–40 adult females from each line in the fifth generation . To determine the proportions of uninserted , R1-inserted and R2-inserted rDNA units , the genomic DNA was digested with ClaI and PstI , fractionated through a 1% agarose gel , and transferred to nitrocellulose . The blot was probed with a segment of the 28S gene ( probe 2 , Figure 1 ) located downstream of the R2 insertion site [29] . Uninserted units gave rise to a 2 . 3 kb ClaI-ClaI fragment , R1-inserted and doubly inserted units to a 1 . 5 kb PstI-ClaI fragment , and R2-inserted units to a 1 . 3 kb PstI-ClaI fragment . To determine the proportions of R1 and R2 double-inserted units , another aliquot of genomic DNA was digested with MspA1I and hybridized with a fragment from the 3′ end of R2 ( probe 3 , Figure 1 ) . The sequence and location of this probe were described by Perez-Gonzalez and Eickbush [10] . Because R2 insertions are located a short distance upstream of R1 insertions , R2-single inserted units gave rise to a 2 . 2 kb hybridizing fragment , while R2 / R1-double inserted units gave rise to a 1 . 3 kb fragment . The size of the rDNA locus was estimated as the number of R2 elements divided by the fraction of the rDNA locus containing R2 insertions . All data on the fraction of rDNA units of each insertion class represents the mean and standard error obtained from four separate blots . Nuclei isolation from 0–23 hour embryos , suspension in 1% InCert agarose , DNA purification , and NotI restriction digestion were conducted as previously described [26] . The digested nuclei plugs were subjected to pulsed-field electrophoresis in 1% agarose at 12°C on a CHEF-DRII apparatus ( Promega ) for 30 hours at 175 volts with switch times of 10 to 30 seconds . The DNA transfer to nitrocellulose and the hybridization a fragment of the 18S rRNA gene were conducted as previously described [26] . | Transposable elements are abundant selfish components of all eukaryotic genomes . Despite the elaborate mechanisms eukaryotes have evolved to control these elements , they continue to proliferate . Here , we study R2 retrotransposons , highly successful elements that only insert into a site within the 28S rRNA genes of animals . In two natural populations of Drosophila , we show that R2 activity is directly linked to the level of R2 transcripts within each fly . The level of R2 transcripts is in turn linked to properties of the tandemly arranged rRNA genes , also known as the nucleolar organizer or rDNA locus . R2 transcript levels appear to depend upon the distribution of R2-inserted units within the rDNA locus , rather than the number of R2 insertions or the total size of the locus . Our findings suggest that R2 transcripts and hence activity can be prevented only when a fly can identify one or more large regions of the rDNA locus free of R2-inserted units . Because recombinations continually change the number and position of rDNA units with R2 insertions , individual loci switch between R2 inactivity and activity , thus perpetuating the long-term survival of R2 elements . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"molecular",
"biology/nucleolus",
"and",
"nuclear",
"bodies",
"genetics",
"and",
"genomics/population",
"genetics",
"genetics",
"and",
"genomics/chromosome",
"biology",
"evolutionary",
"biology/genomics"
]
| 2009 | The Pattern of R2 Retrotransposon Activity in Natural Populations of Drosophila simulans Reflects the Dynamic Nature of the rDNA Locus |
To determine whether treatment of intestinal parasitic infections improves cognitive function in school-aged children , we examined changes in cognitive testscores over 18 months in relation to: ( i ) treatment-related Schistosoma japonicum intensity decline , ( ii ) spontaneous reduction of single soil-transmitted helminth ( STH ) species , and ( iii ) ≥2 STH infections among 253 S . japonicum-infected children . Helminth infections were assessed at baseline and quarterly by the Kato-Katz method . S . japonicum infection was treated at baseline using praziquantel . An intensity-based indicator of lower vs . no change/higher infection was defined separately for each helminth species and joint intensity declines of ≥2 STH species . In addition , S . japonicum infection-free duration was defined in four categories based on time of schistosome re-infection: >18 ( i . e . cured ) , >12 to ≤18 , 6 to ≤12 and ≤6 ( persistently infected ) months . There was no baseline treatment for STHs but their intensity varied possibly due to spontaneous infection clearance/acquisition . Four cognitive tests were administered at baseline , 6 , 12 , and 18 months following S . japonicum treatment: learning and memory domains of Wide Range Assessment of Memory and Learning ( WRAML ) , verbal fluency ( VF ) , and Philippine nonverbal intelligence test ( PNIT ) . Linear regression models were used to relate changes in respective infections to test performance with adjustment for sociodemographic confounders and coincident helminth infections . Children cured ( β = 5 . 8; P = 0 . 02 ) and those schistosome-free for >12 months ( β = 1 . 5; P = 0 . 03 ) scored higher in WRAML memory and VF tests compared to persistently infected children independent of STH infections . A decline vs . no change/increase of any individual STH species ( β:11 . 5–14 . 5; all P<0 . 01 ) and the joint decline of ≥2 STH ( β = 13 . 1; P = 0 . 01 ) species were associated with higher scores in WRAML learning test independent of schistosome infection . Hookworm and Trichuris trichiura declines were independently associated with improvements in WRAML memory scores as was the joint decline in ≥2 STH species . Baseline coinfection by ≥2 STH species was associated with low PNIT scores ( β = −1 . 9; P = 0 . 04 ) . Children cured/S . japonicum-free for >12 months post-treatment and those who experienced declines of ≥2 STH species scored higher in three of four cognitive tests . Our result suggests that sustained deworming and simultaneous control for schistosome and STH infections could improve children's ability to take advantage of educational opportunities in helminth-endemic regions .
Many children in developing countries perform below academically desired levels [1] . Helminth infections are a pervasive part of children's environments in these settings that may contribute to poor educational outcomes through reduced iron status , inflammation , decreased macro-nutrient nutritional status , and distracting symptoms such as abdominal pain [2] , [3] . Some epidemiologic studies have linked these infections to low academic achievement in resource-limited settings [4]–[7] . However , many of the studies did not control for important confounders or had methodological differences that made comparability of findings across studies difficult [8] . All but two prior studies [9] , [10] examined associations between cognitive performance and single helminth species . Recently , polyparasitism , that is , the concurrent multi-species helminth infection , has been associated with childhood anemia and self-reported morbidity [11]–[13] . Its relationship to performance in cognitive tests deserves specific investigation [8] . An earlier cross-sectional study by our group found that moderate or higher intensity infection with Trichuris trichiura , Ascaris lumbricoides , and Schistosoma japonicum were , respectively , associated with low scores on tests of verbal fluency , and the memory and learning subscales of the Wide Range Assessment of Memory and Learning ( WRAML ) tests in school-aged children [14] . It is expected that treatment for parasitic helminth infections will confer a range of benefits to child health , including improvements in academic performance among heavily infected children [15] . However , empirical support for this claim is lacking [8] . Short follow-up periods for most randomized controlled trials , variability in prevalence and baseline intensities of helminth infections , and a background of high re-infection pressure could explain failure to consistently find treatment-associated score improvements . The ambiguity in the literature justifies further exploration of this subject and motivates this longitudinal study to determine the relationship between cognitive testscore improvement and independent declines of schistosome and single soil-transmitted helminth ( STH ) infections , as well as the impact of concurrent declines of two or more STHs on changes in cognitive testscores . Specifically , we provide associations between cognitive testscore improvement and: ( i ) treatment-induced changes in S . japonicum intensity , ( ii ) non-treatment-related or natural declines in single STH infections , and ( iii ) joint infection decline for ≥2 STH species . We hypothesize that no or low level S . japonicum re-infection after praziquantel treatment , and clearance or intensity reductions for single and polyparasitic STH infections will predict improvements in cognitive testscores during follow-up among school-aged children living in a schistosome and STH co-endemic area of Leyte , The Philippines .
The parent study and the nested study reported here were approved by the Brown University , Lifespan , and Philippines Research Institute of Tropical Medicine Institutional Review Boards . Participants' aged ≥18 years provided written informed consent . In addition , all parents/guardians provided written informed consent on behalf of child participants , whereas children aged ≥8 years provided assent . All participants were S . japonicum infected and were treated with the anti-schistosomal drug praziquantel ( 60 mg/kg over 4 hours ) at enrolment as part of the parent study . Only cognitive testing was conducted in a subset of 253 children , aged 7–19 years , as part of this nested observational study . There was no baseline treatment for STH infections as large-scale helminth treatment campaigns were not available in The Philippines at the time this study was conducted . However , at the end of the study , children with STH infection were treated with albendazole and those that became re-infected with S . japonicum were treated with praziquantel . An approach that includes waiting to treat children infected with STH would not be taken today given more recent published findings regarding subtle morbidities related to STH infections . This study was conducted in Macanip , a malaria-free rural rice farming village in Leyte , The Philippines , where S . japonicum and STH infections coexist with high prevalence . This is a nested prospective cohort study conducted in a subset of S . japonicu- infected Filipinos aged 7–30 years who were enrolled in a study of immune correlates of resistance to S . japonicum reinfection [16] . Eligibility criteria included: baseline S . japonicum infection , age 7–19 years at enrolment , provision of parental consent , and child assent for participation in this study . Exclusion criteria included pregnancy or lactation , severe malnutrition ( weight-for-height z-score<−3 ) , severe anemia ( hemoglobin<7 g/dl ) , or the presence of a serious chronic disease determined by history , physical examination , or laboratory findings . Four cognitive tests were administered , including the Philippine nonverbal intelligence test ( PNIT ) , verbal fluency ( VF ) , and two domains of the Wide Range Assessment of Learning and Memory ( WRAML ) , namely verbal memory and learning . Tests were chosen based on their ability to capture a range of cognitive processes including fluid intelligence ( PNIT ) , learning ( WRAML ) , and memory ( VF and WRAML ) while being adaptable across different cultures . The PNIT is an intelligence test that measures concept recognition and abstract thinking [17] . VF test is thought to be a good measure of the central executive component of working memory . The WRAML assesses a child's ability to learn and recall new information . Specifically , the WRAML learning subtests evaluate a child's performance over trials on tasks using the free-recall paradigm , while the WRAML verbal memory subtests assess a child's memory capabilities on meaningful ( i . e . , stories ) and meaningless material ( i . e . , strings of random digits and letters ) [18] . Each of the domains assessed by the WRAML consists of three age-standardized subtests that are added together to derive a total age- and gender-scaled score per domain . Unlike the WRAML , neither the PNIT nor the VF are age standardized; therefore , these tests were adjusted for age variation using linear regression from which we calculated the error terms associated with each child's testscore . We then modeled as the dependent variable the error terms associated with performance in PNIT and VF tests . All tests were translated , adapted for cultural appropriateness , and pilot tested among Filipino children from other S . japonicum-endemic villages near the study area . Testing was conducted in a designated room adjacent to the field laboratory with sufficient lighting and minimal external noises . Ambient temperature within the classroom was approximately 27°C . All children were provided a snack about 30 minutes prior to testing . Joint inter-rater and test-retest reliability with a 6-week interval between tests were evaluated . Cronbach's alpha coefficient was used to assess the degree of internal consistency between tests in the WRAML learning ( α = 0 . 54 ) and WRAML verbal memory ( α = 0 . 81 ) domains . For all tests , higher scores correspond to better performance . Details of each test and its psychometric properties have been previously reported [14] . More details about the rationale for choosing specific tests and their respective properties are presented in Appendices S1 and S2 . Cognitive assessments were made at months 0 , 6 , 12 , and 18 . All infections were assessed at baseline and quarterly thereafter . We have previously reported on cross-sectional associations between helminth infections and performance in the aforementioned tests [14] . Here we determine associations between post-treatment testscores and: ( i ) post-treatment re-infection with S . japonicum and ( ii ) natural infection clearance/decline for STH infections . Only cognitive assessments at 6 , 12 , and 18 months are included in the outcome matrix to preserve temporal sequence between infections and testscore changes . The origin of this prospective analysis is the cohort-wide interval of least infection intensity for all species ( i . e . , months 1–3 ) . STH and schistosome infections were assessed at months 0 , 3 , 6 , 9 , 12 , 15 , and 18 . For S . japonicum only , an additional assessment ( one month post-treatment ) was done to evaluate treatment efficacy . The number of eggs per gram ( EPG ) of stool was determined via duplicate examination of three stool samples by the Kato-Katz method for all species [19] . EPGs were used to define none , low , moderate , or high intensity categories for each species using World Health Organization EPG thresholds [20] . For each individual helminth species , except hookworm , a separate dichotomous baseline intensity indicator was defined as: uninfected/low vs . moderate/high infection to accommodate the intensity distribution in this cohort . For hookworm infection only , baseline infection intensity was defined as none vs . any infection , since >40% of participants were hookworm-free at enrollment and those infected had predominantly low infections . Children were initially grouped by the intensity of concurrent infection with hookworm , A . lumbricoides and T . trichiura as having: ( i ) one or zero low; ( ii ) two or three low; ( iii ) one moderate/high STH; ( iv ) two moderate/high; and ( v ) three moderate/high intensity coinfections [11] . These categories were further combined into one baseline polyparasitic STH indicator to distinguish children with ≥2 STH species at moderate/high intensity ( which may include zero or one low infection of the third STH species ) from those with at most one STH infection at moderate or higher intensity STH coinfection ( other STHs are either absent or present at low intensity only ) . Given our treatment-reinfection design and study inclusion predicated on S . japonicum infection , the most dynamic infection changes occurred with respect to S . japonicum during follow-up; however , STH infection intensity also varied over time . These non-treatment related changes in STH intensity may be due to one or more of the following factors: ( i ) natural changes in STH infections within individuals over time , ( ii ) the limited sensitivity of some STH species to praziquantel [21] , [22] , and ( iii ) lower diagnostic sensitivity for the Kato-Katz method especially when used for the simultaneous assessment of multiple STH species at low intensity in the same host [23] . We defined three post-treatment infection intervals: 1≤t1<6 , 6≤t2≤12 and 12<t3≤18 months; to correspond with the three repeated cognitive assessments . For each STH , t1 infection value ( I1 ) was the mean EPG at month three , whereas for S . japonicum I1 was the mean of EPGs at months one and three . T2 infection ( I2 ) was the mean of EPGs at months six and nine , and t3 infection ( I3 ) was the mean of EPGs at months 12 , 15 , and 18 . Within respective intervals , intra-individual infection change scores ( δit ) were defined by species as follows: t2: δi2 = Ii2 - Ii1; and t3: δi3 = Ii3 - Ii1 . Hence , δit ranged from −∞ to +∞ and will be negative , zero , or positive for a given STH species if the child's infection was lower , equivalent to , or greater than their infection intensity at t1 . For each species , separate δit values were defined and ultimately dichotomized into high vs . low categories as δit≥0 vs δit<0 . For S . japonicum only , infection-free duration was defined as a four level categorical variable that is: ( i ) 0 if not reinfected by month 18; ( ii ) 1 if reinfected between months 12 and 18; ( iii ) 2 if reinfected between months 6 and 12; and ( iv ) 3 if never cured or S . japonicum positive in t1 , t2 , and t3 ( reference group ) . Children reinfected by 6 , 12 , or 18 months were compared to those not reinfected by study end . We determined the number of concurrent STH declines as the sum of individual STH intensity declines using the previously described dichotomous infection decline variable based on δit . Possible values for polyparasitic STH declines were: 0 = no decline/increase STH species , 1 = any one STH , 2 = any two STH to 3 = all STH species intensity decline in a given interval . Using these values , polyparasitic STH decline within intervals was defined as: concurrent intensity decline of ≥2 vs . ≤1 of 3 STH species . We considered an extensive array of potential confounding factors . Because exposure to helminth infection and cognitive testscores vary by age , sex , and socioeconomic status ( SES ) , these factors were considered non-time varying potential confounders . SES measurements were based on baseline questionnaire data addressing four domains of social position; parental and child education , occupation , home/land ownership , and assets . The method used to derive and validate this measure of SES has been described elsewhere [14] , [24] . The derived summary SES variable is divided into four ordinal categories by the quartiles of its distribution . Anemia and nutritional status at baseline were considered potential confounders and/or mediators of low testscores . Anemia was defined on the basis of age- and sex-specific hemoglobin cutoffs recommended by the WHO [25] . Hemoglobin measurement was based on complete hemograms determined on a Serono Baker 9000 hematology analyzer ( Serono Baker Diagnostics , Allentown , PA ) . Nutritional status was assessed using weight-for-age z-scores ( WAZ ) calculated using the National Center for Health Statistics year 2000 reference values in EpiInfo software ( version 2000 , Atlanta , GA ) . Normal and malnutrition status were defined by WAZ≥−2 and WAZ<−2 , respectively . Multivariable random effects regression models were fitted separately to each cognitive test without adjusting for testscore at study enrollment ( month 0 ) given our observational study design [26] . We assumed an unstructured covariance matrix to account for non-independence of repeated cognitive tests within individuals and accounted for clustering of observations within households by including a random intercept for household . Empirical standard errors were used for all estimations to ensure that significance tests were robust against mis-specification of the covariance matrix . In addition , we examined the relationship between test performance and S . japonicum-free duration in separate regression models . Sample regression models for estimation of associations between testscores and S . japonicum infection decline and S . japonicum infection free duration are provided in Appendix S3 . Finally , we examined the potential for modification in the association between infection change and testscore improvement by the following baseline factors: helminth infection intensity , underweight , and anemia . For example , to examine whether the relationship between hookworm infection decline and testscore improvement was heterogenious by hookworm baseline infection intensity , we introduced a three-way multiplicative interaction consisting of the dichotmous indicator of hookworm infection decline , time , and baseline hookworm intensity in a multivariate models that in addition to other confounders also adjust for the baseline intensity of A . lumbricoides , T . trichiura and S . japonicum as well as each of the three dichotmous indicators of change in these infections from the interval of lowest infection . We then examined the p-values associated with interaction terms and where P≤0 . 05 , results are presented by strata of baseline hookworm intensity . The same approach was used to examine baseline underweight and baseline anemia as potential effect modifiers in separate multivariate regression models .
The prevalence of A . lumbricoides , T . trichiura and hookworm infections in this S . japonicum-infectected cohort at baseline were 79 . 9% , 95 . 6% , and 50 . 6% , respectively . Of the 253 children , 97% were concurrently infected by S . japonicum and at least one STH species , approximately 36% were anemic and 60% were underweight relative to U . S . children of the same age and sex ( Table 1 ) . The lowest intensity of S . japonicum infection ( mean = 6 . 8 EPG ) occurred one month post-treatment at which 92% ( n = 217 ) of the sample was infection-free . However , re-infection was rapid and increased steadily until the 12th month of follow-up , at which point 70 . 8% of participants were infected with S . japonicum . Only 25 ( 10 . 6% ) of the re-examined children were free of S . japonicum infection at 18 months . Individual STH intensities also declined from enrollment with the lowest average infection for all STH species occurring at three months . Infection intensity stabilized near this level throughout follow-up for hookworm and T . trichiura infections . The cohort-wide , A . lumbricoides infection intensity by the 18th month was comparable to month zero despite the initial decline post-S . japonicum treatment ( Figure 1 ) . From multivariable models adjusted for sociodemographic characteristics and the intensity of coincident S . japonicum and STH species , declines in the intensity of T . trichiura , hookworm , and polyparasitic STH infections were independently associated with higher average scores on the learning and verbal memory domains of WRAML tests during follow-up . Similarly , A . lumbricoides intensity decline was independently associated with higher scores in the learning sub-scale of WRAML . The intensity of individual infections at enrollment were generally not associated with performance on any of the tests employed , except for moderate/high intensity polyparasitic STH infection , which was associated with lower scores on the PNIT ( Table 2 ) . A decline vs . no change or an increase in S . japonicum intensity from the interval of least infection was not independently associated with improvements in any tests over the study period ( Table 2 ) . We found no evidence that the relationship between S . japonicum infection decline and performance in respective tests differed within strata of S . japonicum intensity at enrollment ( data not shown ) . However children who were S . japonicum free for ≥18 months or those who were S . japonicum infection free until 12 months post-treatment scored higher in all tests relative to rapidly re-infected or persistently infected children . The strength of association was generally attenuated in multi-variable models that controlled for several sociodemographic characteristics and coincident STH and the baseline intensity of S . japonicum infection . Nevertheless , never S . japonicum re-infected children and those S . japonicum infection-free for up to 12 months scored higher in the verbal memory sub-scale of WRAML and VF test , respectively ( Table 3 ) . Anemia and underweight status at enrollment were not independently associated with performance in any tests . However , among children with anemia at enrollment , S . japonicum decline was associated with higher scores on WRAML learning subscale ( mean = 10 . 5 , 95% confidence interval ( CI ) : 4 . 8–16 . 3 ) . There was no association between S . japonicum infection decline and performance in WRAML learning subscale among children without anemia at enrollment ( mean = −3 . 0 , 95% CI: −6 . 4–0 . 4 ) .
Given our observational study design , we cannot exclude residual confounding by unmeasured covariates as an alternative explanation for our findings . By comparing children present at 18-months with those present at baseline on key factors , children scoring in the highest tertile of WRAML verbal memory at baseline and girls were over-represented among those lost to follow-up; however , there was no difference in average hemoglobin , SES , baseline STH intensity and average scores in WRAML learning , PNIT , and VF . In addition , the Kato-Katz relative to other helminth diagnostic methods has been reported to be of lower sensitivity for detecting helminth eggs particularly for individuals with light infections [42] and those with concurrent multi-species infections [23] . We expect that our duplicate assessment of three separate stool samples for each child would have improved the accuracy of helminth diagnosis in this study; however , we are unable to rule out the possible impact of limited sensitivity for lightly infected children . To our knowledge , this is the first longitudinal study to investigate the independent effect of schistosome and individual STH infections as well as that of polyparasitic STH infection decline on learning domains of cognitive function , which may better reflect children's ability to take advantage of limited educational opportunities . The prospective study design , control for coincident helminth infections and numerous other confounders , and the explicit exploration of baseline infection , anemia and nutritional statuses as potential mediators of observed associations are additional strengths of this study . We observed notable fluctuations in T . trichiura and A . lumbricoides intensity in this study even though only S . japonicum infection was treated at enrolment . Praziquantel , however , has been shown to have some anti-hookworm activity [22] . Unlike prior investigations of this question , our analytic strategy highlights the cognitive performance deficits associated with S . japonicum rapid reinfection following treatment as well as the cognitive benefits of natural declines in STH infections among school-aged children . By modeling the relationship between helminth infections and cognitive testscores from the interval of least infection following S . japonicum treatment , we highlight the cognitive test performance advantage of sustained low level single and polyparasitic helminth infections that is derivable in the presence of systematic frequent deworming programs . This relationship may be blunted or lost in an environment characterized by infrequent deworming and high helminth reinfection pressure . Findings from this design and analytic strategy may be more generalizable to the actual implementation of deworming programs than randomized trials . We conclude that declines in the burden of some helminth species and polyparasitic STH infections have beneficial long-term impacts on children's cognitive performance . Our results highlight the benefit of combined control for S . japonicum and STH infections; it further stresses the importance of sustained deworming for improving the learning , memory , and educational attainment of children in helminth-endemic settings . The benefit of combined treatment for these infections notwithstanding , deworming is only a necessary first step in the implementation of a comprehensive integrated helminth control program , which must be tailored to a given endemic setting and include provision of clean water and improved sanitation to mitigate the fundamental causes of these infections and their associated adverse health effects among the most vulnerable populations [43] , [44] . | Parasitic worm infections are associated with cognitive impairment and lower academic achievement for infected relative to uninfected children . However , it is unclear whether curing or reducing worm infection intensity improves child cognitive function . We examined the independent associations between: ( i ) Schistosoma japonicum infection-free duration , ( ii ) declines in single helminth species , and ( iii ) joint declines of ≥2 soil-transmitted helminth ( STH ) infections and improvements in four cognitive tests during18 months of follow-up . Enrolled were schistosome-infected school-aged children among whom coinfection with STH was common . All children were treated for schistosome infection only at enrolment with praziquantel . Children cured or schistosome-free for >12 months scored higher in memory and verbal fluency tests compared to persistently infected children . Likewise , declines of single and polyparasitic STH infections predicted higher scores in three of four tests . We conclude that reducing the intensity of certain helminth species and the frequency of multi-species STH infections may have long-term benefits for affected children's cognitive performance . The rapidity of schistosome re-infection and the ubiquity of concurrent multi-species infection highlight the importance of sustained deworming for both schistosome and STH infections to enhance the learning and educational attainment of children in helminth-endemic settings . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
]
| [
"medicine",
"public",
"health",
"and",
"epidemiology",
"infectious",
"disease",
"epidemiology",
"parasitic",
"diseases",
"helminth",
"infection",
"hookworm",
"infection",
"neglected",
"tropical",
"diseases",
"ascariasis",
"infectious",
"diseases",
"soil-transmitted",
"helminths",
"parasitic",
"intestinal",
"diseases",
"epidemiology",
"hookworm",
"gastrointestinal",
"infections",
"schistosomiasis",
"pediatric",
"epidemiology",
"trichuriasis"
]
| 2012 | Treatment for Schistosoma japonicum, Reduction of Intestinal Parasite Load, and Cognitive Test Score Improvements in School-Aged Children |
To date , no immunization of humans or animals has elicited broadly neutralizing sera able to prevent HIV-1 transmission; however , elicitation of broad and potent heavy chain only antibodies ( HCAb ) has previously been reported in llamas . In this study , the anti-HIV immune responses in immunized llamas were studied via deep sequencing analysis using broadly neutralizing monoclonal HCAbs as a guides . Distinct neutralizing antibody lineages were identified in each animal , including two defined by novel antibodies ( as variable regions called VHH ) identified by robotic screening of over 6000 clones . The combined application of five VHH against viruses from clades A , B , C and CRF_AG resulted in neutralization as potent as any of the VHH individually and a predicted 100% coverage with a median IC50 of 0 . 17 µg/ml for the panel of 60 viruses tested . Molecular analysis of the VHH repertoires of two sets of immunized animals showed that each neutralizing lineage was only observed following immunization , demonstrating that they were elicited de novo . Our results show that immunization can induce potent and broadly neutralizing antibodies in llamas with features similar to human antibodies and provide a framework to analyze the effectiveness of immunization protocols .
HIV transmission remains a huge global public health problem ( www . UNAIDS . org ) . To reduce spread of the virus new prevention methods are being developed based on recent advances in the molecular virology of HIV . In addition to the expanded use of ARVs in new modalities such as pre-exposure prophylaxis and microbicides [1] , this includes commensal microbicides [2] , gene therapies [3] , [4] and vaccines [5] , [6] , [7] . Recent major advances in identifying broad and potent HIV neutralizing monoclonal antibodies ( mAb ) provide invaluable reagents for the development of these strategies to prevent HIV infection . It is well-established that passive infusion of neutralizing mAb can prevent SHIV infection [8] , [9] , [10] , [11] , [12] , [13] and recently it was shown that a mAb can treat infection in non-human primate ( NHP ) models [14] . This combined with the success of antibody-inducing vaccines against other pathogens suggested that a vaccine that can induce neutralizing antibodies at sufficient titers could protect against HIV [15] . While viral escape has not been observed in the NHP models of immunoprophylaxis described above , virus can evolve in response to neutralizing antibodies in HIV-positive patients [16] , [17] and early studies showed escape of patient virus from passive antibody-mediated protection when a single mAb was used [18] . However , recent in vitro work has shown increased neutralization coverage can be achieved by combining the newly identified broadly neutralizing mAb [19] , [20] . These findings highlight the need to ( a ) induce multiple neutralizing antibody lineages for a protective antibody-based vaccine and ( b ) the potential need to use combinations of purified mAb in therapeutic or prophylactic settings . To date , immunizations in human and animal models have yielded antibodies with only limited ability to neutralize HIV [21] , [22] , [23] , [24] , except llama heavy chain only antibodies ( HCAbs ) isolated as individual variable regions ( VHH ) [25] . Previously , we isolated the VHH J3 from a llama immunized with HIV envelope glycoprotein ( Env ) in the form of trimeric recombinant gp140 subunits and found it neutralized 96 of 100 HIV-1 strains by targeting the CD4-binding site ( CD4bs ) [26] . Llamas , along with other camelids have HCAb in additional to conventional heavy and light chain antibodies , [27] . The VHH is the total antigen binding domain or active site of these single-headed immunglobulins ( IgG ) and llama antibodies have been studied in this single domain form due to the dual advantages of bacterial/yeast expression and their high levels of thermal and pH stability [28] . As stable single domains , VHH have advantages for use in microbicides both in terms of gels [28] and protection by commensal bacteria [29] . In this study , we identified three new broadly neutralizing VHH , which bind to the CD4-binding site of the Env subunit gp120 and bind the molecular probe used to isolate VRC01 . Deep sequencing of the VHH phage libraries generated from a set of llamas , which received two different immunization protocols , showed that the new VHH and the previously described anti-HIV VHH J3 [26] were induced by immunization ( Table 1 ) . As such the HIV llama vaccination model is robust and reproducible and demonstrates the potential of a mammalian immune system to produce broadly HIV neutralizing antibodies in response to immunization . We also demonstrate that multiple broadly neutralizing antibody lineages can be raised against HIV in the llama HCAb model and that , when combined as purified VHH , they provide enhanced breadth and potency of neutralization .
Two llamas ( Lama glama ) ( designated Llama 1 and Llama 3 ) were immunized via intramuscular injection of DNA encoding R2 and 96ZM651 . 02 gp160 , and twice more with this DNA in combination with virus-like particles ( VLP ) bearing R2 and 96ZM651 . 02 gp145 protein ( Table S1 in S1 Text ) . Sera taken one week after the last immunization ( t = 54 ) showed measurable binding to homologous gp140 proteins in ELISA but no HIV neutralization activity in the TZM-bl assay ( Figure S1 in S1 Text ) . Both llamas were subsequently immunized a further four times with soluble recombinant gp140 proteins again from R2 and 96ZM651 . 02 HIV strains . In contrast to non-neutralizing response at t = 54 , post immune sera taken from both llamas one week after the final protein immunization ( t = 174 ) neutralized HIV strains from clades A , B and C at 50% inhibitory dilutions ( ID50 ) ranging from 1∶8 up to 1∶200 ( Figure S1 in S1 Text ) . While the potency of these responses is modest , the breadth of activity included ‘difficult-to–neutralize’ strains ( categorized as tier 2 , where tier 1 is easiest and tier 3 is most difficult to neutralize ) [30] from clades A and C , and titers of>1: 100 were seen for both llamas against the homologous strain 96ZM651 . 02 . These serum neutralization titersi were similar to those seen in our previous study where llama 8 was immunized with clade A and B/C gp140 proteins and produced the broadly neutralizing VHH J3 [26] . Thus the serum responses seen were considered strong enough to commence phage library construction and screening for individual VHH using the previously described direct neutralization protocol [26] . VHH clones were initially identified as neutralizing via 384 well robotic screening against two HIV pseudoviruses in the TZM-bl assay ( see methods and materials ) . Initial hits were verified by repeated screening in the TZM-bl assay against a further three viruses . The robotic screening was designed to identify antibodies that neutralize tier 3 strains of HIV against which J3 is either partially or completely inactive by testing initial hits for the ability to neutralize these strains ( Du172 and T257-31 ) at an early time point during the validation process . Three VHH called A14 , B9 and B21 were identified as neutralizing>1 tier 3 HIV pseudovirus and were subsequently purified and IC50 values calculated against sixty one viruses from a range of clades and circulating recombinant forms ( CRF ) ( Table S2 in S1 Text ) . Overall , B9 was the broadest neutralizing VHH , blocking 77% of strains tested with a median 50% inhibitory concentration ( IC50 ) of 0 . 85 µg/ml ( Fig . 1A ) . A14 neutralized 74% of strains tested , but with slightly more potency with a median IC50 of 0 . 53 µg/ml ( Fig . 1B ) . B21 neutralized 72% of strains with a median IC50 of 0 . 8 µg/ml ( Fig . 1C ) . The ability of these VHH to neutralize virus was compared to the previously described VHH J3 and also to another newly identified VHH called 3E3 , which was produced in the same immunization study as J3 but from a different llama ( Table 1 . ) . 3E3 was originally selected from the phagemid library derived from llama 9 by phage display via a competitive elution with soluble CD4 ( sCD4 ) as described previously in [31] . However , re-selection of an identical clone by direct HIV-1 neutralization screening of the llama 9 phagemid library , as described in [26] , indicated the breadth of this VHH and led to its further characterization . 3E3 neutralizes 82% of viral strains from a slightly larger panel of seventy one strains with a median IC50 of 0 . 73 µg/ml ( Fig . 1D ) . However , the ability of 3E3 and the three new VHH to neutralize viruses within individual clades differs from that of J3 , as reflected by the variation in median IC50 values for each clade ( Fig . 1E ) . J3 and 3E3 have similar median IC50 values for each clade , with the exceptions of clade A and CRF AC where 3E3 has lower more potent IC50 values . There is more variation for the three VHH isolated following DNA/VLP/protein immunizations: for clade A and CRF CD , A14 has a log lower median IC50 than J3 , similarly for clades C and CRF CD , A14 is more potent than J3 , but for CRF BC J3 is more potent on average , while for clades G , B and CRF AG , all four VHH have similar median IC50 values . Notably for CRF_AE , A14 , B9 and B21 have IC50 values more than 100-fold lower than J3 . B9 , B21 and A14 all bind in a dose-dependent manner to the gp140 proteins used as the final immunogens ( Fig . 2A , B ) . All three VHH were found to bind to gp140 derived from clade B HIV BX08 ( gp140BX08 ) , as does J3 but not to bind gp41 derived from clade B HIV IIIB ( gp41IIIB ) as does the VHH 2H10 [32] ( Figure S2A , B in S1 Text ) , thus the epitope ( s ) of these three VHH lie within gp120 . 3E3 bound in a dose-dependent manner to the gp140 proteins ( derived from HIV-1 strains 92UG037 and CN54 ) ( Fig . 2C , B ) but also did not bind gp41IIIB . To further pin point where the VHH binds , a gp120 protein derived from the YU2 strain of HIV with a CD4bs mutant unable to bind CD4 ( D368R ) was used . No binding was seen in ELISA to this mutant by 3E3 , A14 , B9 or B21 in comparison to the wild type YU2 gp120 ( Fig . 2E ) . This confirmed that these VHH target the CD4bs of HIV Env as does J3 [26] . However , In contrast to J3 , 3E3 and soluble CD4 , A14 , B9 and B21 all bind to an additional gp120 mutant called RSC3 [33] ( Fig . 2F ) . The RSC3 recombinant gp120 protein was resurfaced to minimize recognition of non-CD4-binding site epitopes and used to probe supernatants of individual B cells from an elite neutralizer patient . This led to the isolation of VRC01 , a mAb that neutralizes 90% of viruses tested [33] . However , neither A14 , B9 nor B21 can bind to the RSCΔ mutant in which residue 371 has been altered to destroy the CD4 binding site ( Fig . 2F ) confirming their interaction with RSC3 is via the CD4-binding site . Crystallographic studies have revealed that VRC01 , and related mAbs , bind to RSC3 by virtue of their angle of approach which is rotated 45° [34] relative to that of CD4 binding observed in a co-crystal with gp120 [35] . Structural studies are needed to assess whether these VHH bind in a similar fashion to VRC01 . However , the total serum response is not dominated by these CD4-specific lineages as illustrated by the unaltered serum binding titres at all time points for both llamas against the RSC3 gp120 protein and the CD4-binding site mutant version RSC3delta ( Figure S1C in S1 Text ) . Analysis of the DNA and encoded amino acid sequences of A14 and B9 revealed that the first two are clonal variants which aligned most closely to the germ line denoted VHH Vg ( T Verrips , unpublished data ) and that the sequence of B21 is highly divergent from both and also aligns to a different germ line V gene sequence Vu . B21 , B9 and A14 belong to two distinct clonal families as compared with the previously described anti–HIV-1 VHH [26] , [31] , [36] , [37] and 3E3 ( Figure S3 in S1 Text ) . Furthermore , there is very limited similarity to the human germ line V gene VH1-2*02 precursor of the VRC01-like broadly neutralizing antibodies isolated from multiple patients . In contrast to the ability of these VHH to bind RSC3 this argues against their being similar to VRC01 . At the nucleotide level the B9 and A14 precursor Vg shares 69 . 83% identity with VH1-2*02 as while B9 and A14 are 68 . 84 and 68 . 73% identical respectively . The B21 precursor shares 45% nucleotide identity with VH1-2*02 and B21 is more dissimilar with only 38 . 75% identity . The closest human V gene to all three VHH precursors is VH3-23*04: B9 and A14 precursor Vg shares 88 . 19% and 82 . 81% nucleotide identity with VH3-23*04 . While the mature B9 and A14 both share 85 . 26% V gene identity with VH3-23*04 and 74 . 51% and 74 . 00% J gene identity with JH5*01 respectively . The B21 precursor Vu shares 78 . 35% with VH3-23*04 but B21 has a much lower identity to any human V gene , the highest being to 81 . 05 to V3-66*02 and sharing 83 . 33% J gene identity with J4*02 . Despite IC50 neutralization potency for some viruses as low as 0 . 02 µg/ml , we hypothesized that the affinity of these VHH could be improved either by additional affinity maturation in the llama or artificially by in vitro mutations of residues predicted to be involved in the interaction with Env . Because there are no known D genes for Lama glama , it has not been possible to identify a germ line sequence corresponding to the CDR3 loop of these antibodies , consequentially nor is it possible to calculate the amount of affinity maturation within this canonical antibody-antigen contact site . However , the presence of two aromatic residues at positions 99 and 105 in both A14 and B9 suggest potential Env contacts . Therefore , both these residues in turn were mutated to glycine to evaluate their relative contributions to the antibody-Env interaction . B9 F99G has a reduced ability to bind gp120 ( Fig . 3A ) and also a log-fold decrease in neutralization potency ( Fig . 3B ) . The W105A mutation in B9 completely removes binding and neutralization function . Thus these residues are preferred and required for function respectively . Given the shared ability of A14/B9 and the VRC01-like antibodies to bind RSC3 , we postulated that the structure-based insertion of an additional aromatic residue prior to CDR2 used to improve the affinity of the VCR01-like NIH45-46 antibody [38] could boost the affinity of A14/B9 . The co-crystal structure of NIH45-46 with gp120 revealed that , despite extensive contacts between the two proteins , the hydrophobic patch on gp120 that is filled by an aromatic residue ( phenylalanine 43 ) when CD4 is bound is not occupied when NIH45-46 binds Env . The insertion of a tryptophan at residue 54 in NIH45-46 resulted in improved affinity and median potency from 0 . 41 µg/ml to 0 . 04 µg/ml and expanded neutralization of six strains of HIV which are resistant to the parental mAb [38] . Thus a tryptophan was inserted into B9 in an attempt to improve its affinity for Env . Two B9 mutants were generated as , due to a longer CDR2 loop in B9 relative to the human antibody NIH45 , it was not clear which residue in B9 corresponds with G54 in NIH45 . Of these two B9 mutants , G53W bound to gp120Bal . 26 with comparable affinity to B9 in ELISA whereas S54W bound more strongly than B9 , with a higher maximum binding level , to an equivalent level of J3 ( Fig . 3A ) . Subsequently , the ability of the mutants to neutralize a pseudovirus bearing the Bal . 26 envelope used in the ELISA was tested . In parallel to its wild type-like ability to bind gp120Bal . 26 , G53W neutralized with a similar IC50 of 2 µg/ml compared to 2 . 2 µg/ml for wild type B9 ( Fig . 3B ) . The other tryptophan insertion pre-CDR2 mutant S54W instead showed increased neutralization potency with a twenty-fold increased IC50 value compared to wild type B9 ( Fig . 3B ) . This shows that the increased affinity seen in the binding studies correlated to improved neutralization against a virus using the same Env protein . In addition , S54W neutralized TRJO4551 . 58 and CH038 strains , which are not neutralized by B9 , with IC50 values of 37 . 3 and 32 . 9 µg/ml . While S54W showed more than a two-fold increase in IC50 against an additional three clade B , two clade C and one CRF BC strains ( Du172 , SS1196 , WITO4160 . 33 , DU156 , CH181 . 12 ) the mutant VHH did not result in improved potency against the other clade B or CRF AG viruses tested ( RHPA4259 . 7 , TR0 . 11 , T266-60 , 236-8 ) ( Fig . 3C , Table S3 in S1 Text ) . Furthermore , there was a trend for enhanced potency for S54W in strains that were less well neutralized by to B9 ( Fig . 3C ) . Over the last 5 years many broad and potent neutralizing mAbs have been identified [20] , [33] , [38] , [39] , [40] , [41] , [42] , [43] , [44] and it has been postulated that combining these mAb could provide protection against a wider range of strains whether by traditional passive transfer at much lower doses than previously imagined [11] or via novel gene transfer strategies [3] . We hypothesized that as human mAb can be successfully combined , without detrimental effects on their individual neutralization activities [19] , A14 , B9 and B21 identified in this study could also be combined with other VHH . These three VHH were mixed with J3 and 3E3 and then assayed for the ability to neutralize HIV in comparison to the same concentration ( 10 µg/ml ) of each VHH in isolation . Viruses with differential susceptibilities to the five VHH were chosen from clades A , B , C and CRF_AG . For all six viruses the mix of five VHH resulted in an IC50 of improved or equivalent potency to the most potent of the five components ( Fig . 4A ) . Therefore , these CD4 binding site VHH do not interfere with one another's function , thus we can extrapolate that the IC50 value for each virus in the panel for the mix of five would be equivalent to that of the most potent VHH against each particular virus ( Fig . 4B ) . This represents an improvement over the use of J3 alone because , while J3 neutralizes 97% of this panel of 61 viruses , the combination neutralizes 100% , and for some viruses the potency of J3 is superseded by A14 , B9 , B21 or 3E3 , resulting in a predicted median IC50 of 0 . 17 µg/ml for the combined antibodies ( Fig . 4C ) . Whether llamas produce neutralizing conventional antibodies against HIV in response to immunization is unknown , but it is plausible that the narrower HCAb are more easily elicited against recessed targets such as the CD4 binding site of HIV than conventional antibodies . Recent data have demonstrated that the anti-HIV potency of J3 is enhanced , when the VHH is presented in a full-length HCAb format , therefore , the neutralization activity of these VHH is not due to their being only the 15 kDa variable region [45] . However , whether or not HCAb can more easily be elicited than IgG does not alter one similarity between neutralizing antibodies derived from HIV-positive patients and J3 , namely that their unmutated ancestors do not bind HIV Env [46] , [47] , [48] , except when modified to do exactly that [49] , [50] , [51] or in one documented case of infection [52] . Reverting just three residues in J3 to germ line removes all ability to bind to either immunogen used to elicit the J3 parental HCAb [26] . Therefore , it is not clear how the antibody was elicited and whether the immunogen interacted with a rare naïve B cell bearing the unmutated ancestor or if the J3 B cell was only boosted by the HIV-based immunizations having been previously affinity matured in response to a different antigen encountered by the field-reared llamas . To resolve this uncertainty , deep sequencing of VHH libraries was performed . This approach has previously provided insight into the development of VRC01-like antibodies in multiple HIV-positive donors [53] , enabled analysis of multiple lineages with distinct specificities arising within an individual donor [54] and allowed in-depth analysis of the heavy chain V gene usage in Env immunized macaques [55] . Briefly , VHH from four immunized llamas and seven naïve llamas were amplified from the phagemid library by PCR using primers specific to the 5′ and 3′ conserved regions of the VHH and subjected to 454 sequencing . Firstly , we sequenced libraries from llamas 8 and 9 from which J3 and 3E3 were isolated respectively . Llamas 8 and 9 were both immunized with gp140 protein derived from HIV strains 92UG037 and CN54 [26] , [31] ( Table 1 ) . Secondly , we sequenced libraries from llamas 1 and 3 described in this study , which produced the VHH A14 , B9 and B21 . Llama 1 and 3 libraries were generated at two separate time points: after the initial three sets of VLP-DNA immunizations and after four sequential gp140 protein boosts . In addition , seven naïve animals previously used to generate naïve VHH libraries from which antigen-specific VHH can be produced via in vitro affinity maturation [56] were sequenced as a control for the variation within the HCAb repertoire of the immunized llamas . All llamas used were genetically outbred and raised outdoors in contrast to laboratory animals often used for immunization studies . Two sequencing runs per VHH library were pooled resulting in 103677 to 213138 unique sequences per sample . Notably , the clonal structure of the VHH repertoire in the naïve animals differed to that in immunized animals ( Table S4 in S1 Text ) . Network diagrams [57] ( Figs . 5 and 6 ) were constructed whereby each point represents an individual sequence Any sequences differing by only one DNA base were connected by linkages . This results in large clusters within a network diagram for large clonal families and many individual points connected to only a few other points representing sequences with fewer clonal relatives . Thus , in a network diagram ( normalized for number of sequences ) with larger clusters , there are more linkages but fewer clonal families representing lower antibody diversity . This scenario is exemplified by the naïve llamas , whose networks had an average of 30269 linkages . In contrast , significantly fewer linkages ( 8843 linkages , p = 0 . 001 ) were seen in the immunized animals and only small clusters , representing an increase in the number of different clonal families as the VHH repertoire diversifies in response to the immunizations ( Figs . 5A , B and 6A , C ) . The mean number of linkages per unique sequence was also found to be significantly larger for naive than immunized llamas , further illustrating the lower degree of divergence between the naive sequences ( naïve: 0 . 22 , immunized: 0 . 052 linkages per unique sequence , p<0 . 005 , Student's t-test ) . The same result is obtained when normalizing the number of unique sequences by total reads per sample ( naïve: 0 . 21 , immunized: 0 . 050 linkages per unique sequence , p<0 . 005 , Student's t-test ) . However , due to the high level of VHH diversity , even in the immunized animals , it is not possible to identify clusters of clones specifically induced by immunization within the network diagrams . This is in contrast to readily identifiable dominant clusters observed in human B cell lymphoma patients [57] where one or few clonal lineages expand massively to dominate the network diagram . Therefore , to establish the relative frequency of neutralizing clones within the VHH repertoire , individual sequences were aligned to both the known neutralizing VHH sequences and their most likely germ line V gene sequences and the percentage identity to the neutralizing VHH plotted ( Fig . 5C , D ) . The percentage nucleotide divergence ( 100 minus the percentage identity ) from germ line and from the respective neutralizing VHH sequence for each of the five neutralizing VHH in each of the 13 llamas was examined . As some reads did not extend to the 5′ end of the V gene , percentage identity was calculated by dividing the total identity by the length of the query sequence . All reads however include the three complementarity determining regions ( CDR ) . Notably , a population of sequences with identity to J3 greater than the germ line precursor ( 85 . 6% ) is found only in immunized llama 8 which produced the J3 HCAb ( population above the dotted line in Fig . 5C right-hand panel compared to naïve llama 3 sequences shown in left-hand panel ) . In addition none of the sequences obtained from any of the naïve llamas are more closely related to J3 than the germ line ( 85 . 6% identity ) ( Figure S4 in S1 Text ) . For the other llama immunized in parallel ( llama 9 ) , the most broadly neutralizing VHH was 3E3 described herein . Again , a population of sequences with identity greater to 3E3 than its germ line precursor Ve ( which differs to that of J3 ) was observed only for llama 9 ( population above the dotted line in Fig . 5D right-hand panel compared to naïve llama 3 in the left-hand panel ) . Interestingly , llama 9 gave rise to no sequences with greater identity to J3 than germ line ( Figure S4 in S1 Text ) and llama 8 gave rise to no sequences closely related to 3E3 . Therefore , the same immunogens and immunization protocol in two different animals gave rise to two separate clonal lineages of CD4-binding site broadly neutralizing HCAbs . Remarkably , both lineages have incurred a three-residue deletion in CDR2 during maturation suggesting the immunogens imposed constraints which resulted in a similar structural solution to high affinity binding that was achieved by different underlying sequences . A limitation of the sequencing analysis of immunized llamas 8 and 9 described above is that they could be compared only to a cohort of seven naïve llamas and not to a pre-immunization time point as no samples from llamas 8 and 9 were available except after the final immunization . In contrast , llamas 1 and 3 immunized in the protocol described herein were sampled at two discrete time points: day 54 after the DNA/VLP immunizations and day 174 after four subsequent gp140 boosts . Serum samples from day 54 showed binding but no detectable neutralization activity ( Figure S1 in S1 Text ) . In contrast , the serum samples from day 174 neutralized pseudoviruses from subtypes A , B and C and the neutralizing VHH A14 , B9 and B21 were isolated from the phage library generated from lymphocytes obtained at day 174 . VHH Networks showed no change in sequence interconnectivity between the two time points in either llama , and both time points showed a higher degree of immunefocusing on the encountered antigen as represented by a lower average number of network linkages as compared to the naïve llamas ( Figs . 5A , B and 6A , C ) . Sequences from both time points were analysed to understand whether clones related to A14 , B9 and B21 arose during the initial DNA/VLP immunization cycle at levels too low to result in sera neutralization and were then boosted by the protein immunizations or whether the protein alone was the antigen responsible for their affinity maturation ( Fig . 6B and D ) . As seen for llamas 8 and 9 , a population of sequences sharing a higher identity with the neutralizing VHH B9 or B21 than their germ lines was seen in both llamas after the protein immunizations at t = 174 . However , no sequences from t = 54 in either llama had greater identity with the affinity matured VHH than the relevant germ line as indicated by the horizontal dotted line on each plot . The same pattern was seen when the closely related A14 VHH was used in place of B9 to analyse the sequences generated from llama 1 and both time points . Thus affinity maturation to a level that produced neutralizing activity was not achieved during DNA/VLP priming phase of the immunizations . Despite representing two separate neutralizing lineages derived from two animals ( Table 1 ) J3 and 3E3 have both incurred a three-residue deletion in CDR2 during maturation and the reversion of this mutation abrogates J3 function [26] . Similarly , re-insertion of the three germ line residues into 3E3 removed its ability to bind gp140 ( Figure S2 in S1 Text ) . This suggests the immunogens imposed constraints resulting in a common structural mode of high affinity encoded by different sequences . To gain insight into which mutations incurred during affinity maturation were responsible for the A14 , B9 and B21 function , germ line reverted VHH ( GL VHH ) were recombinantly produced . These comprised the relevant germ line V gene paired with the mature CDR3 and J sequence of each VHH ( Figure S3 in S1 Text ) . Both GL B9 and A14 showed measurable binding to the R2 gp140 immunogen by ELISA , which was weaker than that seen for the unreverted VHH ( Fig . 2A , B ) , but no binding to 96ZM651 . 20 gp140 ( Fig . 7A , B ) . Thus the key mutation events required for Env binding are within the CDR3 , which in the GL VHH is the same as in the matured VHH . GL B21 on the other hand bound to both immunogens ( Fig . 7A , B ) although to a lesser degree than B21 ( Fig . 2A , B ) which has additional mutations within the V gene . A similar pattern of neutralization activity for the GL VHH was seen , with GL B9 and A14 neutralizing R2 only and GL B21 neutralizing both autologous viruses ( Fig . 7C , D ) although in all cases the neutralization by the GL VHH was less potent than by the mature VHH . This raised the question of whether these GL VHH , which have a mature CDR3 , occurred in the llamas prior to immunization or at t = 54 in addition to at t = 174 when the phagemid library from which the VHH were isolated was generated . Although no neutralization activity was observed at t = 54 it is theoretically possible these VHH could be present but at extremely low frequencies which would not result in neutralizing sera . However , as described above no sequences with greater identity to the affinity matured VHH than germ line were identified at t = 54 ( Fig . 6 ) . To further examine whether any clones similar to the GL VHH produced recombinantly were present at t = 54 additional analysis of the sequencing data was performed . Sequences were filtered into subsets , which shared V gene usage with each VHH . The CDR3 sequences of the subsets were then analyzed for the frequencies of residues that matched to those found in the mature CDR3 at the same position and the number of unique reads where sequential CDR3 residues ( runs ) matched to the mature CDR3 sequence . Total frequencies of matching residues at individual CDR3 positions did not vary substantially between time points for any of the VHH in each llama ( Fig . 7E ) . However , no unique sequences were found at t = 54 which matched the mature CDR3 for more than the initial 5 , 1 , or 3 CDR3 residues for B9 , A14 and B21 respectively ( Fig . 7F , G , H ) following the CAR/CNA residues found at the end of each V gene . In contrast multiple copies of sequences with fully matching CDR3s were found for all VHH in the t = 174 subsets . Thus the GL VHH tested in this study which both bind the R2 immunogen and neutralize the corresponding virus were not present at t = 54 at detectable level by 454 sequencing and it is therefore likely that they were absent or extremely rare at t = 0 .
This study has demonstrated that broadly HIV neutralizing llama HCAb can be reproducibly elicited by immunization ( Fig . 1 ) . Despite diverse immunization protocols , the best HCAb isolated by direct screening of VHH for neutralization breadth all target the CD4-binding site of HIV Env ( Fig . 2 ) and those in most recent immunization study can bind to the CD4-binding site focusing gp120 mutant RSC3 . Notably , the serum binding titres for RSC3 and RSC3delta are highly comparable and thus the sera responses of these llamas are not dominated by RSC3-specific anti-CD4-binding site clones , and the neutralizing VHH described herein are minority variants of the total response , in agreement with the weak neutralization titres observed ( Figure S1 in S1 Text ) and the low overall frequency of sequences belonging to these lineages ( Fig . 6 ) . It may be that only CD4-binding site specific lineages were isolated due to easier elicitation of such specificities due to the site's conservation within the unstabilised gp140 protein immunogens used ( unlike trimer-specific epitopes ) or due to the preference of the single-headed HCAbs for long CDR3 protrusions which can effectively bind into recessed canyons such as the CD4-binding site . It must also be noted that only one broadly neutralizing lineage in each animal was identified from 5000 clones and other specificities may have been found by screening more clones . In addition , although the screening was performed on phagemid libraries derived from llama PBMCs using primers designed to amplify all known VHH families ( T Verrips , unpublished data ) bias may have been introduced by this process if certain sequences were less preferentially incorporated in to the library . Similarly , some VHH may not have been well expressed in the bacterial system , preventing them from being identified by the robotic screening . Interestingly the fine specificities of CD4 binding vary between clones resulting in the ability of each clonal lineage of VHH to effectively neutralize some strains of HIV which are resistant to the other VHH ( Fig . 1 ) . A14 , B9 and B21 are highly potent against representatives from HIV strains circulating in certain geographical areas that J3 does not target as efficiently . We have demonstrated that combining five of these VHH results in superior neutralization breadth compared to that of any of the VHH used in isolation ( Fig . 4 ) . This concurs with studies of human antibodies targeting independent epitopes [19] . However , all five VHH target the CD4 binding site , nevertheless , they can be used in combination to result in neutralization potency equivalent that of to the most potent VHH . This suggests the differences in affinity of each VHH for any particular virus are great enough to prevent the less potent VHH occupying the CD4 binding site resulting in less potent neutralization . That VHH elicited by immunizations which target the same site on Env can be used in combination without compromising neutralization potency is encouraging for vaccine design studies . It indicates that the co-induction of a variety of antibodies targeting the same site is not detrimental to neutralization function in itself , although how this can be achieved by immunization remains to be determined . Since the combination of these five anti-CD4 binding site VHH neutralized 100% of strains tested with an IC50 of less than 0 . 2 µg/ml ( Fig . 4 ) , their combined use in a topical anti-HIV microbicide [28] could provide a higher barrier to infection and viral escape than the use of any one VHH individually . It has previously been shown that virus escape occurs from passively transferred human antibodies used to treat infection when three or less are used in combination whereas no escape was seen from a combination of five antibodies [58] . How significant escape would be in the setting of chronic microbicide use remains to be determined but the use of multiple VHH targeting the same site in different ways should reduce the risk and provide a higher barrier to transmission . We have also shown that the VHH described can be improved in terms of potency and breadth via site-directed mutagenesis . The residues altered were selected based on the observation that the three VHH described herein , B9 , A14 and B21 , bind to the RSC3 probe that was used to isolate the VRC01-class of human CD4 binding site broadly neutralizing antibodies ( Fig . 2 ) and that previously the introduction of an aromatic residue into the VRC01-like antibody NIH45-46 resulted in increased breadth and potency [38] . The introduction of an aromatic in the mutant B9S54W also resulted in increased potency and breadth ( Fig . 3 ) despite the lack of significant sequence homology between VRC01-like antibodies and B9 , A14 and B21 ( compounded by the lack of a light chain which contributes to the VRC01-Env interface ) . Thus , it cannot be concluded that these VHH are VRC01-like or that these immunogens can elicit such antibodies in humans . It would be of interest to undertake structural studies to establish whether there are any structural similarities . While the overall frequency of A14 , B9 and B21-like sequences is low there is a clear expansion of these lineages after the animals received four gp140 protein boosts ( Fig . 6 ) . This low frequency agrees with the weak neutralization breadth seen for the t = 174 serum samples ( Figure S1 in S1 Text ) . In contrast , the deep sequencing analysis of VHH present at t = 54 shows no over-representation of the sequences which share higher levels of identity with the neutralizing VHH than the germ line ( Fig . 6 ) . This suggests that the low-level maturation achieved during the DNA/VLP immunizations was not sufficient for neutralization or the definition of these clonal lineages . Interestingly , A14 , B9 and B21 GL VHH ( with the V gene reverted to germ line but mature DJ regions ) were all able to bind the R2 immunogen and neutralize the homologous virus ( Fig . 7 ) but only B21 GL could bind the both immunogens and neutralize both viruses . This difference in binding ability to the two immunogens suggests the R2 immunogen may have driven the earlier stages of affinity maturation of this lineage in llama 1 , which resulted in clones that could also recognize 96ZM651 . 02 gp140 as the immune response progressed and the lineage developed breadth via mutations in the V gene region . That these germ line reverted VHH could bind and neutralize at first appears to contradict the lack of neutralization activity and the inability to identify clones belonging to each lineage at t = 54 . However , further sequencing analysis established that these germ line chimera were not present at the earlier time point , as no sequences with the fully mature DJ regions paired to the relevant germ line V genes were detected at t = 54 whereas between 50 and 250 copies of such clones were present at t = 174 . Thus we can conclude that mutations occurred in the CDR3s during the protein immunization phase resulting in neutralization ability , alongside additional mutations in the V gene which are required for breadth at least in the A14/B9 lineage . The germ line chimeras do however demonstrate that these VHH require only mutations within their CDR3 for neutralization activity . This is in agreement with the finding that the F99G and W105G substitutions in the CDR3 of B9 diminish and destroy its anti-HIV activity respectively ( Fig . 3A , B ) . This is in contrast with the fact that J3 and 3E3 require a deletion within the V gene which shortens the CDR2 loop by 3 residues [26] , ( Figure S2 in S1 Text ) to bind and neutralize HIV Env . Notably , the global incidence of A14 , B9 and B21-like sequences is low within even the t = 174 , however this is across total VHH from whole peripheral blood , not just those specific for these immunogens because no panning enrichment was performed with the phagemid library prior to neutralization screening . This greatly increases the signal-to-noise ratio in the deep sequencing analysis and thus it is not possible to identify which clones are immunogen-specific and analyse these as a subset . This is because a primary sequence alone does not allow us to predict function unless the sequence is highly similar to a characterized clone . Therefore , future studies should involve an immunogen-specific flow cytometry selection of the immungen-specific llama B cells prior to generating the library to more easily identify clusters of immunogen-specific VHH . Such studies may also provide insight into whether the high level of divergence from germ line ( relative to human immunization studies ) seen in these clones is standard in the HCAb response to immunization . That HCAb memory B cells may be more mutated from germ line relative to conventional IgG B cells is not entirely unanticipated . Naïve HCAbs inherently start from a less diverse paratope repertoire as they do not combine a heavy and light chain , both derived from VDJ recombination . Therefore , in order to successfully advance through successive rounds of affinity maturation and compete for immunogen with the conventional antibodies , llamas also produce ( estimated at 70% of total IgG ) the HCAbs that may undergo additional mutations . In addition , the mutation process is inherently random , and deleterious ( stop codons etc . ) . In a normal B cell a deleterious mutation in either chain will prevent expansion of that lineage , whereas in a HCAb B cell a successful heavy chain does not require a functioning light chain to proceed . It is significant that parallel immunization of llamas 8 and 9 gave rise to two separate clonal lineages of CD4 binding site broadly neutralizing HCAbs which have both incurred a three residue deletion in CDR2 during maturation . This finding indicates that the immunogens imposed constraints , which resulted in a similar structural solution to high affinity binding achieved by different sequences . This has implications for the use of deep sequencing analysis of immunization studies: if the sequence of J3 had been used as a reference to filter the sequences from llama 9 ( without prior screening of VHH from llama 9 ) , no J3-like antibodies would have been found , 3E3 would not have been identified and it would have falsely appeared that the immunization of llama 9 had failed . These findings are consistent with the observation that HIV Env immunization activates a highly polyclonal B cell response of substantial genetic diversity in NHP [55] . This is in contrast to the well-characterized VRC01-like family of CD4-binding site antibodies identified in multiple HIV-positive donors , which share both a heavy chain V gene precursor and unusual features at a DNA sequence level which could allow sequencing-based identification of other similar antibodies in distinct patients [53] . Thus , this study has implications for analysis of human vaccine studies , as in addition to searching for defined lineages it is worthwhile to perform functional analysis of monoclonal antibodies that may have found new structural solutions to high affinity binding which cannot be discerned from DNA sequence alone . A critical implication of this work for the field of HIV vaccine design is the observation that the most potent and broad individual anti-HIV VHH , J3 , was elicited in response to the gp140 immunogens used . It has been suggested that the extensive affinity maturation of antibodies and HCAbs which neutralize HIV could be the result of prior affinity maturation of clones to a non-HIV antigen . Hypothetically , such a B cell clonal lineage could have a greater affinity for Env than any germ line B cell and could either be boosted or further affinity matured in response to Env immunization . Deep sequencing analysis has shown that this is highly unlikely in the case of J3 as no sequences were generated by deep sequencing of the seven naïve llamas or three other immunized llamas that share greater identity with J3 than its putative germ line V gene . Paradoxically , it remains the case that even minimal mutation towards the J3 germ line V gene abolishes the ability to bind either immunogen used [26]; the reintroduction of the three germ line CDR2 residues renders the VHH incapable of binding Env . However , it must be noted that the llama immunization model is not only an animal model but one resulting in a subtype of antibodies not produced in humans . Furthermore , the neutralizing HCAb induced occurs in the llamas at a low frequency that does not result in broadly neutralizing sera , the goal for a protective HIV vaccine . However , this model has allowed us to examine four HIV broadly neutralizing clonal lineages induced by vaccination , which has not been possible in other animal models to date , and highlights the many challenges of evaluating immunization studies with deep sequencing of antibody variable regions .
This study was performed in strict accordance with the Dutch Experiments on Animals Act 1997 . In accordance with article 18 of the Act , the protocol was assessed and approved by the Animal Ethics Committee of Utrecht University ( permit number: DEC#2007 . III . 01 . 013 ) . All efforts were made to minimize discomfort related to immunizations and blood sampling . The animal welfare officers of Utrecht University checked the mandatory administration and supervised the proper conduct of procedures and the well-being of the llamas that were used . Llamas 1 and 3 were immunized via intramuscular injection of plasmid DNA encoding R2 and 96ZM gp160 in PBS ( 7 . 5 mg of DNA ) with or without virus like particles bearing R2 and 96ZM envelope proteins in PBS ( protein content 50 µg ) . Subsequently additional immunizations were administered with intramuscular injection of ZM96 & R2 gp140 protein ( 50 µg each ) in a freshly prepared 4 . 5-ml water in oil emulsion prepared by vigorously mixing 2 volumes of antigen with 2 . 5 volumes of the adjuvant Stimune ( CEDI Diagnostics ) . Immunizations and VHH library construction were performed as described previously ( De Haard et al . , 2005 ) . In brief , the llamas received one dose of DNA alone , then two doses of DNA combined with VLP , followed by four doses of protein in adjuvant as per Table S1 in S1 Text . The anti-Env immune response in sera was verified via neutralization of three viruses in TZM-bl cells ( Figure S1 in S1 Text ) . Total RNA was isolated from between 120 and 150 million peripheral blood lymphocytes ( PBMC ) collected after immunization ( on day 54 and 174 ) and cDNA was prepared . The VHH repertoire was amplified and cloned into the pCAD50 phagemid vector . To obtain recombinant bacteriophages expressing the VHH as fusion proteins with the M13 bacteriophage gene III product , transformed TG1 E . coli cells were grown to logarithmic phase and then infected with helper phage M13KO7 . The phage particles were precipitated with polyethylene glycol . Codon optimized and c-terminal truncated ( aa714–856 , HXB2 numbering ) R2 and 96ZM651 gp145 genes were synthesized ( GeneArt ) and cloned into pcDNA3 . 1 using NheI and PmeI restriction sites and NheI and XhoI respectively . DNA for immunizations was prepared using the Qiagen EndoFree Plasmid Giga Kit according to the manufacturer's instructions . A Giga-prep was performed to obtain at least 60 mg of DNA . 200 µg of R2 virus like particles ( VLPs ) and 200 µg of 96ZM VLPs were made . Pseudotyped VLPs for immunization purposes were produced in 293F cells by transient transfected with a codon-optimized , Rev-independent gene for Gag ( IIIB ) [59] and the respective envelope encoding plasmid in a ratio of 2∶1 . VLPs were harvested 72 h post-transfection , cleared by centrifugation at 3000 g for 15 min , loaded onto a 30% sucrose in PBS cushion ( 5 ml for 30 ml of supernatant ) and ultra-centrifuged at 100 , 000 g for 2 h . The pellet was resuspended overnight in PBS and stored at −80°C . For the quantification of incorporated envelope protein an ELISA was used . Each subsequent ELISA washing step was carried out with 200 µl PBS+0 . 05% Tween20 and each incubation was done for one hour at room temperature . A Clear 96-well MaxiSorp plates ( NUNC ) were coated overnight with 1 µg/ml of antibody 5F3 ( Polymun ) . Plates were blocked using 10% fetal calf sera in PBS+0 . 05% Tween20 followed by three washings . Pseudotyped VLPs were denatured in the presence of 0 . 5% Triton-X for 1 h and the VLPs as well as recombinant gp140 standard protein were added in serial dilution followed by incubation . After three additional washing steps 50 µl of a 1∶1000 dilution of antibody MH23 ( NIBSC ) was added and incubated . The plate was washed three times before adding 0 . 65 µg/ml anti-mouse horseradish peroxidase–conjugated Ab ( Dako ) followed by incubation . After six washes TMB ELISA substrate was added , and the plates were incubated until standard proteins were visible . Purified gp140 from R2 and from 96ZM were mixed with Stimune commercially available Stimune adjuvant ( CEDI Diagnostics , Lelystad , The Netherlands ) . Recombinant trimeric gp140 from HIV-1 92UG037 ( clade A ) for ELISAs was provided by S . Jeffs ( Imperial College London , London , England , UK ) . Recombinant D368R and wild-type monomeric gp120 from HIV-1 YU2 ( clade B ) and recombinant RSC3 and RSC3Δ37 for ELISAs were provided by J . Mascola ( National Institutes of Health [NIH] , Bethesda , MD ) . Recombinant gp41 from HIV-1 IIIB , recombinant trimeric gp140 from HIV-1 CN54 and BX08 were obtained from the CFAR , NIBSC and were donated by Immunodiagnostics and Polymun Scientific , respectively . TZM-bl cells ( Derdeyn et al . , 2000; Wei et al . , 2002; Li et al . , 2005 ) were obtained through the NIH AIDS Research and Reference Reagent Program from J . C . Kappes ( University of Alabama at Birmingham , Birmingham , AL ) , X . Wu ( NIAID , NIH ) , and Tranzyme , Inc . and cultured in Dulbecco's modified Eagle medium ( Invitrogen ) containing 10% ( vol/vol ) FCS . Pseudoviruses were generated from two separate plasmids , one encoding a full length HIV virus with a defective env and the other encoding a functional env . This results in non-replication competent HIV progeny viruses that undergo only one cycle of infection , which is sufficient to test the ability of an antibody to inhibit HIV entry into cells . HIV-1 IIIB was propagated in H9 cells all other replication-competent virus stocks were prepared from HIV-1 molecular clones by transfection of 293T cells . HIV-1 IIIB ( ARP101 ) was obtained from the CFAR , NIBSC . IIIB was donated by R . Gallo and M . Popovic ( University of Maryland School of Medicine , Baltimore , MD ) . HIV-1 Env pseudotyped viruses were produced in 293T cells by co-transfection with the pSG3Δenv plasmid ( Kirchherr et al . , 2007 ) . The subtype B and C HIV-1 Reference Panels of Env Clones ( Li et al . , 2005 , 2006 ) were obtained through the NIH AIDS Research and Reference Reagent Program , Division of AIDS , National Institute of Allergy and Infectious Diseases ( NIAID ) , NIH . HIV-1 subtype CRF07_BC Gp160 clones , subtype CRF02_AG Gp160 clones ( 263-8 , T278-50 , and T266-60 ) , and the 92UG037 , 93MW965 . 26 , and 96ZM651 . 02 Gp160 clones were provided by D . Montefiori ( Duke University Medical Center , Durham , NC ) through the Comprehensive Antibody Vaccine Immune Monitoring Consortium ( CA2 VIMC ) as part of the CAVD . All additional pseudoviruses were produced at the CAVD Preclinical Neutralizing Antibody Core laboratory . Serum samples were heat-inactivated to destroy complement by incubation at 56°C for 1 h before use in neutralization assays . Threefold serial dilutions of llama sera were then tested , starting at a 1∶5 dilution in the 96-well plate the TZM-bl cell-based assay developed by Derdeyn et al . ( 2000 ) , Wei et al . ( 2002 ) , and Li et al . ( 2005 ) , with Bright-Glo luciferase reagent ( Promega ) using a Pherastar plate reader ( BMG Labtech ) . Phages expressing the cloned VHH repertoire were plated onto agar containing 100 µg/ml ampicillin and 2% syncytial stain ( 1 g methylene blue and 0 . 33 g basic fuchsin in 200 ml methanol ) . Individual clones were picked using a Norgren CP7200 colony picker ( RapidPick; Hudson Robotics ) into 384-well master plates . 6144 individual clones were expressed in TG1 E . coli cells in a 384-well plate format . Each clone was expressed in 150 µl of 2× TY medium containing 100 µg/ml ampicillin and 0 . 1% glucose , followed by induction of VHH production with 0 . 1 mM isopropyl-β-dthiogalactopyranoside . Bacterial pellets were frozen at -80°C for a minimum of 1 h and then thawed and resuspended in PBS . The periplasmic extract from each well was separated from bacterial debris by filtration through a 0 . 45-µM polyvinylidene fluoride membrane and screened for the ability to neutralize HIV-1 . To enable high-throughput screening and characterization of VHH , neutralization was measured using 50 50% tissue culture infective doses of virus in a 384-well plate adaption of the 96-well plate the TZM-bl cell-based assay developed by Derdeyn et al . ( 2000 ) , Wei et al . ( 2002 ) , and Li et al . ( 2005 ) , with Bright-Glo luciferase reagent ( Promega ) using a Pherastar plate reader ( BMG Labtech ) . DNA from the individual VHH that neutralized any tier 2/3 viruses to <20% seen with control was purified , sequenced , and recloned into the pCAD51 expression vector followed by transformation into TG1 cells for purification and further characterization . Expression from the pCAD51 vector incorporates a 6-His and a c-Myc tag to the C terminus of the VHH and removes the bacteriophage gene III product . The VHH were purified by means of the attached His tag using TALON Metal Affinity Resin ( Takara Bio Inc . ) . Mutagenesis of VHH DNA was achieved using the QuikChange site-directed mutagenesis kit ( Agilent Technologies ) according to the manufacturer's instructions . Mutant VHH protein was expressed and purified as described above . The neutralization activity of the VHH was assayed in duplicate/triplicate at either University College London or VIMC laboratories . No virus inactivation was observed with a negative control VHH with no specificity for HIV was used as control ( De Haard et al . 2005 ) . or with a pseudovirus bearing a mouse leukemia virus Env . VHH IC50 titers were calculated using the XLFit4 software ( IDBS ) or the Labkey Neutralizing Antibody Tool ( Piehler et al . , 2011 ) . Cran R radial plots were used to display the inverse IC50 values ( Fig . 1 ) . Clear 96-well MaxiSorp plates ( Thermo Fisher Scientific ) were coated overnight with 2 µg/ml of recombinant Env . Plates were blocked using 5% milk powder in TBS . Serial dilutions in TBS supplemented with 0 . 05% Tween ( TBS-T ) containing 1% milk powder ( TMT ) of the VHH to be assayed and of a negative control VHH were then added to the plates in triplicate wells , and the plates were incubated at room temperature for 1 h and subsequently washed four times with TBS-T . The wells were then incubated with 0 . 5 µg/ml mouse anti–c-Myc–horseradish peroxidase–conjugated Ab ( Roche ) in TMT for 1 h at room temperature . After six washes with TBS-T , TMB ELISA substrate ( Thermo Fisher Scientific ) was added , and the plates were incubated at 37°C for 0 . 5 h . Absorbance at 450 nm was detected , and background-subtracted data were plotted against VHH concentration . Two technical replicates were run for each sample and the corresponding sequence datasets were merged . Any reads that were identical sub-sequences of other reads were removed and the read count for the longer sequences adjusted accordingly . Raw MiSeq reads were filtered for base quality ( median>32 ) using the QUASR program ( http://sourceforge . net/projects/quasr/ ) . Overlap between forward and reverse reads , where no nucleotide mismatches are allowed . MiSeq forward and reverse reads were merged together if they contained identical overlapping region of>65 bp , or otherwise discarded . No nucleotide mismatches are allowed in the overlap regions , as this would be indicative of sequencing error or recombination . Primer sequences were trimmed from the reads , and sequences retained for analysis only if both primer sequences were identified with 100% match . Non-immunoglobulin sequences were removed and only reads with significant similarity to reference Llama IgHV and IgHJ genes using BLAST ( Altschul et al . , 1990 ) were retained ( 10-10 , 10-3 e-value threshold respectively due to differences in gene length ) . Reads were retained if they contained complete open reading frames ( without stop codons ) . Non-functional BCRs , PCR error or recombination may lead to the artificial introduction of stop codons . Length filter: between 150 and 320 nucleotides . If recombination was to occur , this filter would remove reads that had significant changes in their length . Graphs for each sample were generated using the igraph R library [60] with nodes corresponding to individual read sequences and linkages between nodes whose sequences differed by at most one nucleotide . Node sizes were proportional to the number of observed reads for each sequence . Sequence similarities were calculated by performing BLASTX [61] searches on read sequences against a database of known V gene products and reference sequences ( J3 , 3E3 , A14 , B9 and B21 ) , counting insertions and deletions in the alignments as non-identical residues . Divergence was calculated as ( 100 – percentage sequence similarity ) . Divergence plots were generated using the ggplot2 R library [62] . Custom Perl and R scripts were used throughout to parse and analyse the sequence datasets ( R Development Core Team , 2008 http://www . R-project . org ) . | Developing a vaccine against HIV-1 is a priority , but it remains unclear whether immunizations in humans can elicit potent broadly neutralizing antibodies able to prevent HIV-1 transmission . Llamas possess heavy chain only antibodies and conventional heavy and light chain antibodies . We previously reported the heavy chain only antibody J3 , which potently neutralizes more than 95% of HIV strains , and was induced by immunization . Here we immunized two further llamas and elicited three novel broadly neutralizing heavy chain only antibodies , which were identified by high-throughput screening . These neutralizing llama antibodies target different areas of the CD4-binding site of the virus , therefore breadth and potency are increased when they are used in combination . To gain greater understanding of how the llama immunizations worked , deep sequencing of the HIV binding region of the antibodies was performed . This revealed that the antibodies were matured fully only in response to the protein immunogens . Furthermore , the VHH elicited in different animals , while sharing functional hallmarks , were encoded by distinct sequences and thus could not have been identified by a deep sequencing analysis alone . Our results show that immunization can potentially induce protective antibodies in llamas and provide a method to more extensively evaluate immunization studies . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"organisms",
"immunodeficiency",
"viruses",
"viruses",
"infectious",
"disease",
"immunology",
"clinical",
"immunology",
"medical",
"microbiology",
"viral",
"pathogens",
"microbial",
"pathogens",
"biology",
"and",
"life",
"sciences",
"immunology",
"microbiology",
"vaccination",
"and",
"immunization",
"hiv-1",
"immune",
"response"
]
| 2014 | Molecular Evolution of Broadly Neutralizing Llama Antibodies to the CD4-Binding Site of HIV-1 |
Cells can maintain their functions despite fluctuations in intracellular parameters , such as protein activities and gene expression levels . This commonly observed biological property of cells is called robustness . On the other hand , these parameters have different limitations , each reflecting the property of the subsystem containing the parameter . The budding yeast cell cycle is quite fragile upon overexpression of CDC14 , but is robust upon overexpression of ESP1 . The gene products of both CDC14 and ESP1 are regulated by 1∶1 binding with their inhibitors ( Net1 and Pds1 ) , and a mathematical model predicts the extreme fragility of the cell cycle upon overexpression of CDC14 and ESP1 caused by dosage imbalance between these genes . However , it has not been experimentally shown that dosage imbalance causes fragility of the cell cycle . In this study , we measured the quantitative genetic interactions of these genes by performing combinatorial “genetic tug-of-war” experiments . We first showed experimental evidence that dosage imbalance between CDC14 and NET1 causes fragility . We also showed that fragility arising from dosage imbalance between ESP1 and PDS1 is masked by CDH1 and CLB2 . The masking function of CLB2 was stabilization of Pds1 by its phosphorylation . We finally modified Chen's model according to our findings . We thus propose that dosage imbalance causes fragility in biological systems .
Intracellular biochemical parameters , such as gene expression levels and protein activities , are highly optimized in order to maximize the performance of biological systems [1]–[4] . On the other hand , these parameters operate within certain limitations to maintain the function of the system against perturbations such as environmental changes , mutations , and noise in biochemical reactions . This robustness against fluctuations in parameters is considered a common design principle of biological systems [5]–[7] . The cell cycle is a series of events that leads to cellular duplication , and the regulatory system is highly sophisticated to precisely maintain cellular integrity [8] . The budding yeast Saccharomyces cerevisiae is an excellent model organism to understand the principle of cell cycle regulation because of its ease in use with molecular genetic techniques . Cell cycle regulation has been integrated into a mathematical model called Chen's model [9] . This model implements interactions of about 25 genes involved in the budding yeast cell cycle to reproduce over 100 mutant phenotypes , and thus , has become a standard for measuring the robustness of the budding yeast cell cycle [10]–[12] . The robustness of a cellular system can be assessed by perturbation analysis of the extent to which each intracellular parameter can be changed without disrupting the function of the system [1] . To assess the robustness of the budding yeast cell cycle , we used a previously developed genetic experiment designated “genetic tug-of-war” ( gTOW ) to measure the copy number limit of overexpression of certain target genes [13] . In gTOW , a target gene with its native promoter is cloned into a special plasmid , and the plasmid copy number can be increased just before cell death ( Figure 1 ) [13] . In this method , the copy number limit of gene overexpression is measured as a fold increase and compared with its native expression level . Using gTOW , we measured the copy number limit of overexpression of 30 cell cycle-related genes that varied from <2 to >100 [13] . Although these numbers are thought to reflect the robustness of the subsystems harboring these genes , it is not easy to identify the molecular mechanism behind the phenomenon causing the variation because robustness arises from interactions between multiple components of the system . Analysis using mathematical models helps to identify the mechanism responsible for the robustness of biological systems [1] , [14] . We compared the gTOW data with Chen's model and discussed the mechanisms underlying fragility and robustness of the yeast cell cycle in response to overexpression of several genes [13] . In this study , we define a cellular system has robustness if its normal mode of operation is hardly destroyed even when amount of a certain component in the system largely fluctuates . And we define a cellular system has fragility if its normal mode of operation is easily destroyed when amount of a certain component fluctuates . In this study , “fluctuation of component” corresponds to the increase of gene copy number in the cell , and the increase of gene expression parameter in the computer simulation ( both manipulations cause gene overexpression ) . When the cell is viable despite overexpression of a certain gene , we call that the cellular system has robustness upon overexpression of the gene . When the cell is not viable due to minor overexpression of a certain gene , we call that the cellular system has fragility upon overexpression of the gene . We observed that the copy number limit for a mitotic phosphatase gene CDC14 overexpression was very low ( <2 ) , which was well predicted by Chen's model ( <2 ) . In contrast , we observed that the copy number limit for the separase gene ESP1 overexpression was quite high ( >160 ) , and the prediction of Chen's model ( <1 . 4 ) was quite different from the upper limit in vivo [13] . According to multiple reports [15]–[18] , in Chen's model , enzymes such as Cdc14 phosphatase and separase are regulated by direct 1∶1 binding with their inhibitors ( Net1 and Pds1 ) ( Figure 2A and 2C ) . And overexpression of CDC14 cured the lethality brought about by overexpression of NET1 [18] . We thus predict that fragilities upon overexpression of these genes arise from dosage imbalance between the enzymes and their inhibitors [13] . However , it has not been shown that dosage imbalance between CDC14 and NET1 causes fragility of the yeast cell cycle . Moreover , there is a discrepancy between predictions of the model and the experimental data in case of the copy number limit of ESP1 as mentioned above . In this study , we analyzed the molecular mechanisms underlying fragility of CDC14 regulation and robustness of ESP1 regulation . On the basis of our observations , we suggest that dosage imbalance between enzyme and its inhibitor causes cellular fragility . We further suggest that knowledge about cellular robustness can be effectively used to improve integrative mathematical models .
To experimentally determine whether dosage imbalance causes fragility of the yeast cell cycle upon overexpression of certain genes , we first thoroughly analyzed the quantitative relationship between CDC14 and NET1 , as well as ESP1 and PDS1 using Chen's model . When the regulated enzyme ( i . e . , Cdc14 or Esp1 ) alone is overexpressed and its amount exceeds that of its inhibitor ( i . e . , Net1 or Pds1 ) , the cell cycle halts due to abnormal chromosome separation ( Figure S2A and S2C ) . When the amount of enzyme and its inhibitor increased simultaneously , the cell cycle proceeds normally ( Figure S2B and S2D ) . We performed two parameter viability tests in which parameters for expression of enzymes ( Cdc14 and Esp1 ) and their inhibitors ( Net1 and Pds1 ) were systematically increased and the ability of the cell cycle to persist with any combination of these parameters was tested . Computational analysis showed that to maintain the cell cycle , the absolute amount of enzymes and their inhibitors did not matter , but the quantitative ratio was important and needs to be conserved [ ( fold increase in expression of NET1 ) / ( fold increase in expression of CDC14 ) = 0 . 95–1 . 95; ( fold increase in expression of PDS1 ) / ( fold increase in expression of ESP1 ) = 0 . 56–2 . 20] ( Figure 2B and 2D ) . If fragility upon overexpression of CDC14 is caused by dosage imbalance against NET1 , this conserved ratio should be observed in vivo as well . Similar to computational analysis , we designed an experiment by adding another multicopy plasmid carrying NET1 into the gTOW experiment of CDC14 ( Figure 3A ) . This experiment called “2D-gTOW” is based on the fact that multicopy plasmids with CDC14 or NET1 replicate with 2µDNA origin , exist as multi-copy in a cell and their numbers vary among the cellular population [19] . Moreover , the copy number of the gTOW plasmid can be controlled by changing leucine concentration in growth media; the average plasmid copy number within a cell is around 35 in the presence of leucine but increases to more than 150 in the absence of leucine ( Figure 1 ) [13] . As expected , introduction of NET1 plasmid prevented cellular death upon overexpression of CDC14 ( Figure 3B ) . In the rescued cells , the average plasmid copy number of CDC14 increased dramatically ( ∼40 copies per cell; Figure 3C ) , and the amount of Cdc14 protein also increased accordingly ( Figure 3D ) . We then performed the two parameter viability test by measuring the copy numbers of the plasmids in multiple independent experiments with and without leucine in medium . Most importantly , the ratio between CDC14 and NET1 was clearly conserved [ ( NET1 copy number ) / ( CDC14 copy number ) = 1 . 38 , R2 = 0 . 96]] ( Figure 3E ) , similar to that observed in computational analysis . Moreover , when we use cdc14-1 , a temperature-sensitive CDC14 gene with reduced activity , the ratio was reduced but still conserved [ ( NET1 copy number ) / ( cdc14-1 copy number ) = 0 . 77 , R2 = 0 . 94] ( Figure 3F ) . We show for the first time that cellular fragility upon overexpression of CDC14 is caused by dosage imbalance between NET1 and CDC14 . While there is evidence that Esp1 is regulated by 1∶1 binding of Pds1 [15] , [16] , cellular fragility from dosage imbalance has not been observed ( i . e . , the copy number limit of ESP1 overexpression is high ) . We thus hypothesized that there was an additional regulatory mechanism besides simple binding of Pds1 [13] . To demonstrate ESP1 regulation by other factors , we performed another 2D-gTOW experiment in various gene knockout mutants ( Figure 4A ) . Among 23 nonessential cell cycle gene knockouts , cdh1Δ and clb2Δ strain showed significant reduction in the copy number limit of ESP1 ( Figure 4B ) . The fragility of these knockouts upon overexpression of ESP1 was rescued by additional PDS1 plasmids ( Figure 4C ) . Moreover , the ratio between ESP1 and PDS1 copy number was well conserved in cdh1Δ cells [ ( PDS1 copy number ) / ( ESP1 copy number ) = 1 . 27 , R2 = 0 . 91] ( Figure 4D ) , as observed between CDC14 and NET1 . This result indicates that dosage imbalance between ESP1 and PDS1 actually causes cellular fragility upon overexpression of ESP1 , but additional regulations by CDH1 and CLB2 mask the potential fragility . Our study above indicated the existence of some factors regulating ESP1 that are not incorporated into Chen's model . We thus tried to improve Chen's model by seeking additional regulations to reproduce the gTOW results . We should note that recently , a study published more detailed model for M-phase-specific regulation [20] . Although this model implements additional regulations such as signaling activity of Esp1 toward FEAR ( Cdc14 early anaphase release ) pathway ( see below ) , the model still predicted fragility upon overexpression of ESP1 ( Figure S3 ) , indicating the regulation we are seeking is not implemented in this model . We focused on regulation by Clb2 , a B-type cyclin-dependent kinase ( B-CDK ) subunit , because there is evidence that B-CDK is involved in ESP1 regulation . In the budding yeast , B-CDK phosphorylates the inhibitor Pds1 to regulate its localization [15] and stability in metaphase [21] . In higher eukaryotes , CDK phosphorylates separase ( Esp1 homolog ) to inhibit its protease activity [22] . However , it has never been shown whether any of these regulations confer cellular robustness upon overexpression of Esp1 . Therefore , we first modified Chen's model by incorporating each regulation into three independent computational models and tested if they gave viable solution with; overexpression of ESP1 , overexpression of ESP1 in the absence of Clb2 , and simultaneous overexpression of ESP1 and PDS1 ( Figure 5A and Table 1 ) ( details in Text S1 , S2 , S3 , S4 , S5 , S6; Table S1 , S2 , S3; and Figure S1 , S4 , S5 , S6 , S7 , S8 ) . Among them , the models for Esp1 phosphorylation and Pds1 stabilization could well reproduce the behaviors of the cell in terms of copy number limits of ESP1 overexpression ( Table 1 ) . We next experimentally verified these regulations limiting the ESP1 copy number . When phosphorylation by Clb2 is involved in cellular robustness upon overexpression of ESP1 , regulation can be destroyed by introducing mutations in phosphorylation sites of the target proteins . In Esp1 of the budding yeast , putative CDK phosphorylation sites [ ( Thr/Ser ) -Pro] are observed , which are conserved among relative yeast species ( Thr-1012 , Ser-1025 , and Thr-1032 ) ( Figure 6A ) . We substituted these amino acids with alanine ( esp1-AAA ) , and measured the copy number limit by gTOW to verify the ESP1 phosphorylation model , and found that the limit was >100 ( Figure 6B ) . This indicates that direct phosphorylation of Esp1 by Clb2 does not confer robustness upon overexpression of ESP1 . A study reported that phosphorylation of Pds1 on Thr-27 and Ser-71 by B-CDK stabilizes Pds1 , and its regulation is required for synchronization of chromosomal partition [21] ( Figure 5A , Pds1 stabilization model ) . We thus built our model according to their findings , and our model predicted that the copy number limit of ESP1 was significantly reduced when phosphorylation of Pds1 was inhibited ( Figure 5D ) . We then measured the limit of ESP1 in the alanine-substituted mutants on these phosphorylation sites ( pds1-2A ) , and found that the cell did not accept the high copy number of ESP1 as observed in clb2Δ cells and the limit of overexpression was significantly decreased ( Figure 6C ) . This is the first evidence to show that Pds1 phosphorylation is involved in cellular robustness upon overexpression of ESP1 . We should note that the decrease of the limit of ESP1 overexpression in pds1-2A cells was not dramatic as in clb2Δ cell ( Figure 7 ) , suggesting that there is another mechanism by which clb2 confers cellular robustness against ESP1 overexpression . One important but not reported assumption to build the Pds1 stabilization model was that the amount of Pds1 is in large excess of Esp1 ( Pds1∶Esp1 is 112∶1 on average during the cell cycle , see Text S1 ) , while their amount is almost the same in Chen's model ( 0 . 998∶1 ) . To confirm our assumption , we measured the quantitative ratio of Pds1∶Esp1 using TAP-tagged proteins . Although we could not detect Esp1 expressed from the chromosomal copy , the amount of Pds1 was at leaset more than 64-fold higher than Esp1 ( Figure 6D and 6E ) , supporting our assumption . We thus conclude that quantitative regulation of Pds1 through phosphorylation by B-CDK requires for masking the fragility arising from dosage imbalance between Esp1 and Pds1 . Esp1 is known to have two independent activities . One is a protease activity to digest certain substrates such as Scc1 and Slk19 [23] , [24] , and the other is a signaling function to activate FEAR pathway that is a pathway to activate Cdc14 [20] , [25] . We thus tested if either of these activities was the determinant of limit of Esp1 overexpression in cdh1Δ and clb2Δ strains . We measured the limit of esp1-C1531A , an ESP1 allele without separase activity [26] in the wild type , cdh1Δ , clb2Δ , and pds1-2A strains respectively . The limit of esp1-C1531A overexpression was increased in the clb2Δ strain and the pds1-2A strain up to >100 copies ( Figure 7 ) . Interestingly , however , the limit of esp1-C1531A overexpression was still very low in the cdh1Δ strain ( Figure 7 ) . These results indicate that CLB2 and CDH1 are involved in the robustness of ESP1 regulation in different ways; CLB2 is involved in the regulation associated with the protease activity , and CDH1 is involved in the regulation associated with the FEAR signaling activity .
Knowing the mechanisms causing cellular fragility is important for controlling cellular functions or finding novel drug targets [27] , [28] . In this study , we demonstrated that dosage imbalance between Cdc14 and Net1 causes significant cellular fragility upon overexpression of CDC14 using computational and experimental analysis . We believe that 2D-gTOW can be used as an experimental technique to detect cellular fragility arising from dosage imbalance . As in one of the examples , we were able to detect potential fragility arising from dosage imbalance between ESP1 and PDS1 , although it was masked by CDH1 and CLB2 . Using this method , we would be able to show more examples of dosage imbalances causing cellular fragility . Because the strain having mutations on the phosphorylation sites of Pds1 by Clb2 did not accept Esp1 overexpression ( Figure 6C ) , we concluded that the masking function of Clb2 is performed through the stabilization of Pds1 . On the other hand , currently we could not explain the masking function of CDH1 . The function of CDH1 in Esp1 regulation is at least different from CLB2 , because the limit of overexpression of the esp1-C1531A mutant was still low in the cdh1Δ strain , but high in the clb2Δ strain ( Figure 7 ) . This fact suggests that the masking function of CDH1 is performed through the process downstream of FEAR pathway , which regulates the activity of Cdc14 phosphatase [25] . Cdh1 is a component of ubiquitin-conjugating enzyme complex called APC that degrades a number of proteins such as Clb2 , Cdc5 , Cdc20 , Cin8 , etc . [29]–[33] . CDH1 will thus confer the robustness of Esp1 regulation through degradation of these M-phase components . One possible function of Cdh1 to confer cellular robustness against the overexpression of Esp1 is performed thorough a polo-like kinase Cdc5 , which also regulates Cdc14 activity [34] . When Cdh1 is inactivated , the substrate Cdc5 activity will increase , and Cdc14 will be activated . In the situation , the cell will be very sensitive against further activation of Cdc14 by the FEAR pathway due to the overexpression of Esp1 . Alternatively , the masking function of Cdh1 could be performed through Cdc20 . Cdc20 is another component of APC , which promotes the degradation of Pds1 [35] . Because Cdc20 is a potential target of Cdh1 [32] , [33] , the activity of Cdc20 could be higher in the cdh1Δ strain . It is thus possible that the amount of Pds1 is reduced in the cdh1Δ strain due to the over-activation of Cdc20 , which causes reduction of the robustness of ESP1 regulation . In addition to the mechanisms described above , there could be other mechanisms that make the cellular system robust against Esp1 overexpression . For example , Pds1 is considered as a chaperone for Esp1 [36] , [37] , which will make excess Esp1 over Pds1 unstable . Although we did not adopt the Pds1 transport model ( Figure 6 ) to explain our finding , it is also a quite effective mechanism to regulate the activity of Esp1 . CDC55 , a component of PP2A phosphatase and a direct regulator and a downstream effector of Esp1 [20] , [38] ) , will be also involved in the robustness . M-phase regulations with components such as Cdc5 and Cdc55 should be implemented into the integrated model , and verified further combinational gTOW experiments to uncover the whole regulatory mechanisms conferring the cellular robustness against Esp1 overexpression . We should note that the reason why the clb2Δ cell and cdh1Δ cell are fragile against overexpression of ESP1 , could be arisen from the same mechanistic consequence as the observation that clb2Δ and cdh1Δ are synthetic lethal with pds1Δ [39] , [40] , although Pds1 phosphorylation by Clb2 should be an exception . Unfortunately , Chen's model and our modified model do not reproduce the behavior of pds1Δ mutant ( i . e . , viable in real cell , but essential in the models [9] , data not shown ) . We thus could not test these phenotypes in our model . We hope that modifications of the model by implementing the regulation above will solve the discrepancy . In the budding yeast cell , there are several genes , such as actin encoding gene ( ACT1 ) or beta-tubulin-encoding gene ( TUB2 ) , that cause extreme fragility due to imbalance against binding partners [41] , [42] . Dosage balance ( i . e . , stoichiometry ) between histone dimmer sets must be conserved for normal mitotic chromosome transmission [43] . We thus hypothesize that dosage imbalance is a common cause of cellular fragility . In regulation of CDC14 , dosage imbalance is exposed whereas in regulation of ESP1 , it is masked . In many cellular processes , it is likely that fragilities caused by regulation through 1∶1 binding ( here we call “stoichiometric regulation” ) will be masked . In the case of Esp1 , what we found here ( and Chen's model did not implement ) was that the inhibitor Pds1 was in large excess of the separase Esp1 ( Figure 6 ) . Excess of the inhibitor could be a general mechanism by which the systems are robust against dosage fluctuation of the enzyme . In case of Cdc14 and Net1 , the amount of both proteins within the cell are the same order ( Net1-TAP exists with 1 . 59E+03 molecules/cell and Ccd14-TAP exists with 8 . 55E+03 molecules/cell ) [44] , this is one of the reasons of the exposed fragility . However , as a trade off , the excess inhibitor should be effectively and timely inactivated when activation of the enzyme is required . Separase needs to be activated accurately in the period of metaphase to anaphase transition . Phosphorylation of Pds1 on Thr-27 and Ser-71 by Clb2 is the regulation that makes the precise inactivation ( degradation ) of Pds1 , which requires the cell cycle system to be robust against overexpression of Esp1 . Regulations conferring cellular robustness might therefore be generally discovered around stoichiometric regulations , as was observed in case of ESP1 . Moreover , we may be able to control cellular robustness by modifying the regulators around stoichiometric regulation . How is fragile regulation advantageous for a cell ? Regulation by simple protein-protein interactions is one of the simplest ways to generate ultrasensitive responses in cellular systems [45] , [46] , and might have evolved to add novel regulations toward enzymes . For example , multiple CDK inhibitors are present in yeasts to mammalian cells , but they are quite diverse . While B-type cyclins Clb2 ( S . cerevisiae ) and Cdc13 ( S . pombe ) are quite similar ( BLAST E-value 6e-79 ) , their inhibitors Sic1 and Rum1 do not show any similarity ( BLAST E-value >0 . 05 ) . This suggests that these factors have evolved independently from different ancestor proteins to achieve the common purpose of binding and inhibiting CDK . In addition , drugs for molecular targeted therapy utilize the mechanism of stoichiometric regulation against the target . This is the only known enzymatic regulation thus far that humans can design . In fission yeast and higher eukaryotes , no stoichiometric regulator for Cdc14 phosphatase homologue is known to exist [47] . We propose that during evolution , the budding yeast uniquely acquired Cdc14 regulation with Net1 , but it conversely produced fragility caused by dosage imbalance as a trade-off . The activity of Cdc14 itself is quite tightly regulated by two signalling pathway designated FEAR and MEN ( mitotic exit network ) , which are found only in the budding yeast [48] . The budding yeast may have uniquely acquired these regulations in order to buffer the fragility due to the dosage imbalance . Developing integrative cellular models with high predictive ability is one of the goals of systems biology . However , it is sometimes criticized that large-scale integrative cellular models are indefinitely adjustable and can no longer be proven false [49] . For this purpose , a general experimental technique to effectively evaluate and refine models is needed . In this study , we evaluated a model with data for cellular robustness obtained by gTOW , found discrepancies , modified them according to the current knowledge for reproducing robustness , and evaluated them with combinatorial gTOW . We believe that this analytical scheme will be effective for further development of integrative cellular models .
A wild-type yeast strain BY4741 ( MATa , his3Δ1 , leu2Δ0 , met15Δ0 , ura3Δ0 ) and its derivatives with deletion of cell-cycle-related genes ( in Figure 4 ) were obtained from Open Biosystems Inc . Haploid yeast strains KK001 ( leu2Δ , ura3Δ , PDS1 ) and KK002 ( leu2Δ , ura3Δ , pds1-T27A , S71A ) are progenies of LH651 and LH557 [21] , respectively . To detect TAP-tagged Pds1 and Esp1 , derivatives of a yeast strain SC0000 ( MATa , ade2 , arg4 , leu2-3 , 112 , trp1-289 , ura3-52 ) , SC4998 ( PDS1-TAP-klURA3 ) and SC1033 ( ESP1-TAP-klURA3 ) ( Euroscarf ) were used . Yeast cells were cultured in synthetic complete medium ( SC ) lacking indicated amino acids . SC medium was prepared using YNB with ammonium sulfate ( MP Biomedicals , LLC ) with DO supplement ( Clontech ) and 2% glucose . Plasmids used in this study are listed in Table 2 . pTOWug2 is a pSBI40 derivative carrying URA3-GFP fusion gene instead of URA3 . pRS423-mRFP is a pRS423 derivative carrying HIS3-RFP fusion gene instead of HIS3 . gTOW experiments were performed as described previously [13] . For 2D-gTOW , cells transformed with both pSBI40 and pRS423 derivatives were cultivated in SC without uracil and histidine , and then they were transferred into SC without uracil , histidine , and leucine . The copy numbers of pSBI40 and pRS423 derivatives were measured using real-time PCR as described previously [13] , except that an HIS3 primer set ( HIS3-1F , TTCCGGCTGGTCGCTAAT and HIS3-1R , GCGCAAATCCTGATCCAAAC ) was used to measure the copy number of pRS423 derivatives . Data shown in Figure 3C , Figure 4B and 4C , Figure 6B and 6C , and Figure 7 are averages of at least four independent experiments . Cdc14 and Net1 proteins were quantified by western blot analysis using their specific antibodies ( sc12045 and sc27758; Santa Cruz Biotechnology , Inc . ) as described previously [13] . TAP-tagged proteins were detected using PAP ( P1901l; Sigma-Aldrich ) . Numerical simulations were carried out using Matlab version 7 . 3 . 0 . Chen's model and Queralt's model were implemented on Matlab script files [9] , [20] . The code for Chen's model was based on that obtained from Dr . Cross . For details and codes used in this study refer to Text S1 . | Normal cell functioning is dependent on balance between protein interactions and gene regulations . Although the balance is often perturbed by environmental changes , mutations , and noise in biochemical reactions , cellular systems can maintain their function despite these perturbations . This property of cells , called robustness , is now considered to be a design principle of biological systems and has become a central theme for systems biology . We previously developed an experimental method designated “genetic tug-of-war , ” in which we assessed the robustness of cellular systems upon overexpression of certain genes , especially that of the budding yeast cell cycle . Although the yeast cell cycle can be maintained despite significant overexpression of most genes within the system , the cell cycle halts upon just two-fold overexpression of M phase phosphatase CDC14 . In this study , we experimentally showed that this fragility is caused by dosage imbalance between CDC14 and NET1 . Interestingly , fragility of regulation of separase gene ESP1 , potentially caused by dosage imbalance , was masked by regulation of other factors such as CDH1 and CLB2 . We thus propose that dosage imbalance causes fragility in biological systems . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"cell",
"biology/cell",
"growth",
"and",
"division",
"computational",
"biology/systems",
"biology",
"biochemistry/theory",
"and",
"simulation",
"computational",
"biology/molecular",
"genetics"
]
| 2010 | Fragilities Caused by Dosage Imbalance in Regulation of the Budding Yeast Cell Cycle |
The Escherichia coli curved DNA binding protein A ( CbpA ) is a poorly characterised nucleoid associated factor and co-chaperone . It is expressed at high levels as cells enter stationary phase . Using genetics , biochemistry , and genomics , we have examined regulation of , and DNA binding by , CbpA . We show that Fis , the dominant growth-phase nucleoid protein , prevents CbpA expression in growing cells . Regulation by Fis involves an unusual “insulation” mechanism . Thus , Fis protects cbpA from the effects of a distal promoter , located in an adjacent gene . In stationary phase , when Fis levels are low , CbpA binds the E . coli chromosome with a preference for the intrinsically curved Ter macrodomain . Disruption of the cbpA gene prompts dramatic changes in DNA topology . Thus , our work identifies a novel role for Fis and incorporates CbpA into the growing network of factors that mediate bacterial chromosome structure .
Bacterial chromosomes are organised into a nucleoid by an integrated network of supercoiling , transcription and nucleoid associated DNA binding proteins [1] . This network is highly dynamic and responsive to changes in the extracellular environment . In Escherichia coli , a particularly notable change in nucleoid structure occurs when cells are starved . Namely , the nucleoid adopts a super compact conformation that is believed to protect the genome . These changes in nucleoid structure coincide with changes in supercoiling [2] , transcription [3] and the available pool of nucleoid proteins [4] . Strikingly , only two nucleoid proteins are specifically up-regulated as cells approach stationary phase; the DNA binding protein from starved cells ( Dps ) and Curved DNA binding protein A ( CbpA ) [4] . The Dps protein has been studied for decades and has well characterised DNA binding , compaction and protection properties [5]–[7] . In growing cells expression of Dps is blocked by Fis , the major growth phase nucleoid protein [8] . In sharp contrast to Dps , the regulation and DNA binding properties of CbpA have hardly been studied . The CbpA protein was first isolated as a factor present in crude E . coli cell extracts that preferentially bound curved DNA in vitro [9] . The affinity of CbpA for DNA is similar to that observed for other nucleoid associated proteins [10] . CbpA consists of an N-terminal J-domain separated from two C-terminal domains ( CTDI and II ) by a flexible linker [11] . The J-domain interacts with the modulator protein CbpM , which inhibits CbpA co-chaperone activity and DNA recognition [12]–[14] . DNA binding activity locates to the linker-CTDI region and CTDII mediates dimerisation [11] , [15] . Transcription of cbpA initiates from overlapping promoters referred to as P1 and P2 [16] . Most cbpA transcription is driven by the σ38 dependent P2 promoter with σ70 dependent P1 making only a small contribution [16] . Consistent with this , CbpA accumulates in starved E . coli , reaching 15 , 000 copies per cell after two days [4] . In contrast to Dps , which is uniformly distributed within the nucleoid , CbpA forms nucleoid associated foci [17] . Nothing is known about the function of CbpA in starved E . coli cells . In this work we sought to better understand the regulation and function of CbpA . We show that Fis plays a crucial role in preventing CbpA expression in growing cells . Hence , Fis binds to DNA target sites in the cbpA regulatory region . When bound to these sites Fis prevents cbpA being transcribed from an aberrant promoter , located within the coding sequence of an adjacent gene . In starved cells , when cbpA transcription is induced by σ38 , CbpA binds pervasively across the E . coli genome with a bias to the intrinsically curved Ter macrodomain . Disruption of the cbpA gene prompts dramatic changes in DNA supercoiling .
As E . coli cells approach stationary phase they induce expression of two nucleoid proteins , Dps and CbpA . Previously , we found that Dps expression in growing E . coli cells is repressed by Fis [8] . Thus , we wondered if Fis may also control production of CbpA . In an initial experiment we cloned a 302 base pair DNA fragment , containing the entire cbpA regulatory region and part of the adjacent yccE gene , upstream of lacZ ( illustrated schematically in Figure 1Ai ) . We then measured LacZ activity in WT JCB387 cells , carrying this fusion , throughout growth . We observed basal levels of LacZ activity until the onset of stationary phase at which point LacZ activity increased ( green line in Figure 1Aii ) . Strikingly , in JCB3871Δfis cells , LacZ activity was high throughout the time course ( red line in Figure 1Aii ) . In complementary western blotting experiments we measured Fis levels in E . coli cells at different stages of growth ( Figure 1Aiii ) . Fis levels were inversely correlated with cbpA induction . Intrigued by this observation we measured Fis binding to the cbpA regulatory DNA using Electrophoretic Mobility Shift Assays ( EMSA ) . For comparison , we also tested DNA fragments that drive expression of other nucleoid proteins and the nirB promoter , which has a high affinity for Fis [18] . The raw EMSA data are shown in Figure S1 and a quantification of the experiment is shown in Figure 1B . The data show that Fis binds particularly tightly to the cbpA regulatory region . Note that Fis forms three distinct complexes with the cbpA regulatory DNA , suggesting three separate Fis binding sites ( Figure S1 ) . Fis binds to a 15-base-pair AT-rich DNA target that is highly degenerate . A common feature of many Fis sites is a G at position 1 and a C at position 15; however even these features are not universally conserved [19]–[21] . Thus , DNAse I footprinting was used to locate Fis bound at the cbpA regulatory region . The full sequence of the 302 base pair DNA fragment is shown in Figure S2 . The previously characterised cbpA P1 and P2 promoters are highlighted . Throughout this work , all numbering is with respect to the P1 promoter . The results show that Fis binds a DNA element between 90 and 145 base pairs upstream of the cbpA P1 transcript start ( Figure 2A ) . Since Fis binds to a 15 base pair recognition sequence the large 55 base pair footprint most likely contains three Fis binding sites , as observed in our EMSA analysis ( Figure S1 ) . Scrutiny of the DNA sequence corresponding to the region bound by Fis identified one match to the canonical Fis binding sequence ( i . e . containing a G at position 1 and a C at position 15 ) . This site is centred 101 base pairs upstream of the P1 transcription start ( Figure S2 ) and can be disrupted by altering the key positions in the Fis recognition sequence . Thus , the cbpA-108C-94G DNA fragment has a greatly reduced affinity for Fis in EMSA assays ( Figure S3 ) . We presume that adjacent DNA sites for Fis in this region must have an atypical sequence . This is not exceptional and similar observations have been made for Fis binding sites at the rrn promoters [20] . As a first step to understanding control of cbpA by Fis we utilised KMnO4 footprinting that detects DNA melting around transcription start sites . Thus we were able to measure RNA polymerase binding in vitro , to the cbpA regulatory region , in the presence and absence of Fis . Recall that cbpA transcription can be stimulated by σ70 or σ38 associated RNA polymerase . The results in the absence of Fis , illustrated in Figure 2B , show different patterns of σ38 ( lane 2 ) and σ70 ( lane 5 ) dependent DNA opening . As expected , our analysis identified DNA melting at the known P1 and P2 promoters ( highlighted by red and blue boxes in Figure 2B ) . Surprisingly , we also observed σ70 dependent DNA opening at three further locations ( highlighted by green , orange and purple boxes in Figure 2B ) and one additional σ38 dependent DNA opening event ( highlighted by a black box in Figure 2B ) . Given the large effect of Fis on cbpA expression in vivo ( Figure 1Aii ) , we were surprised that addition of Fis to the KMnO4 footprinting reaction had minor effects ( lanes 3–4 and 6–7 ) . Thus , Fis only inhibited DNA untwisting at position −100 ( open complex 4 ) which overlaps the Fis binding element . To aid interpretation of the in vitro KMnO4 footprinting analysis we conducted in vivo mRNA primer extension experiments . This enabled us to identify cbpA transcript start sites , used in the presence and absence of Fis , in growing and stationary phase cells . The results of this analysis are shown in Figure 2C . In WT cells we observed only two cbpA mRNA primer extension products . As expected these correspond to the P1 and P2 promoters ( see lanes 1 and 3 in Figure 2C ) . The Δfis mutation had little effect on transcription start site selection in stationary phase cells ( see lane 4 in Figure 2C ) . However , the Δfis mutation had a dramatic effect in growing cells ( see lane 2 in Figure 2C ) . Thus , in growing Δfis cells , the most abundant cbpA transcript did not originate from either the P1 or P2 promoter . Rather , it initiated form a site located 204 base pairs upstream of the P1 promoter in the yccE gene ( highlighted in purple in Figure 2C ) . This transcription start site , labelled P6 in the schematic , aligns precisely with open complex 6 in the KMnO4 footprinting experiments ( compare Figure 2B and 2C ) . An additional transcription start site , which we refer to as P4 , aligns with open complex 4 ( see orange highlighting in Figure 2B and 2C ) . Note that additional primer extension products in lane 2 of Figure 2C , which are also close to open complex 4 , are likely due to degradation of longer transcripts . To confirm that we correctly identified the P4 promoter , we created a P4::lacZ fusion and investigated the effect of mutating the proposed −10 element ( Figure S4A ) . The data show that the P4 promoter drives LacZ expression and that transcription from the P4 promoter increases when the proposed −10 element is improved ( Figure S4B ) . However , overall P4 makes only a small contribution to cbpA transcription . To investigate further the effect of Fis on each cbpA promoter we used in vitro transcription assays . The 302 base pair cbpA regulatory DNA fragment was cloned upstream of the factor-independent λoop transcription terminator in plasmid pSR . The different mRNA products expected to be produced using this DNA template are illustrated in Figure 3Ai . We also created two derivatives of this construct lacking either the P6 promoter or a combination of the P6 , P2 and P1 promoters . The different promoters were disrupted by making mutations in the promoter −10 element . Thus , the −217G–216G mutation negates the P6 promoter , the −11G mutation disrupts the P2 promoter and the −7G–6G mutation abolishes transcription from P1 ( Figure 3Ai ) . Figure 3Aii shows the results of in vitro transcription assays with the different DNA templates . As expected , using the wild type DNA template , a mixture of σ38 and σ70 associated RNA polymerase stimulated production of transcripts corresponding to the P1 , P2 , P4 and P6 promoters ( Figure 3Aii lane 1 ) . The −216G–217G mutation abolished transcription from the P6 promoter ( Figure 3Aii lane 2 ) . Similarly , mutations −6G , −7G and −11G , which disrupt the P1 and P2–10 hexamers respectively , abolish transcription from the P1 and P2 promoters ( Figure 3Aii lane 3 ) . Note that , the P1 and P2 transcripts differ in length by only 5 nucleotides and cannot be resolved in this experiment . Lanes 4 and 5 of Figure 3Aii show the pattern of transcripts generated in the absence and presence of Fis respectively . As expected , Fis greatly reduced transcription from the P6 and P4 promoters . Conversely , Fis had only a minor effect on the P1 and P2 promoters . Note that the small RNAI transcript , from the pSR replication origin , acts as an internal control . Taken together our data suggest that Fis prevents CbpA expression in growing cells primarily by silencing the strong P6 promoter . To confirm this we designed cbpA::lacZ fusions lacking either P1 and P2 , or P6 , due to mutations in the promoter −10 hexamer . Our hypothesis was that disruption of the P6 promoter would negate the effect of Fis whereas disruption of P1 and P2 would not . The different DNA fragments , fused to lacZ , are illustrated in Figure 3B alongside LacZ activity data from growing JCB387 or JCB3871Δfis cells . As expected , LacZ expression from the wild type cbpA::lacZ fusion increased in JCB3871Δfis cells . An almost identical result was obtained using the −11G , −7G , −6G fragment lacking P1 and P2 . Hence , Fis does not exert its effect via the P1 and P2 promoters . The DNA fragment lacking the P6 promoter , due to the −217G and −216G mutations , stimulated low levels of LacZ expression that did not increase in JCB3871Δfis cells . To confirm that the Fis binding region was responsible for mediating repression of the P6 promoter we generated a series of P6::lacZ fusions containing nested deletions downstream of the P6 transcription start site . The different DNA fragments , fused to lacZ , are illustrated in Figure 4 alongside LacZ activity data from growing JCB387 or JCB3871Δfis cells . The starting P6::lacZ fusion , containing the P6 promoter and full Fis binding region , stimulated lacZ expression that increased in cells lacking fis ( Figure 4A ) . Deletions of between 10 and 30 base pairs ( Δ10 , Δ20 and Δ30 ) sequentially place the Fis binding element closer to the P6 promoter . These deletions had little effect on either LacZ expression or repression by Fis ( Figure 4Bi ) . Conversely , larger 60 , 80 and 100 base pair deletions ( Δ60 , Δ80 , Δ100 ) , which sequentially degrade the Fis binding element , lead to stepwise increases in LacZ expression and a reduction in repression by Fis ( Figure 4Bii ) . Thus , our data are consistent with Fis “protecting” cbpA from the effects of the P6 promoter by interacting with the Fis binding region . We next sought support for our model by re-evaluating published data from three independent chromatin immunoprecipitation ( ChIP ) studies of Fis binding across the E . coli chromosome [22]–[24] . We also scrutinised our own unpublished mRNA deep sequencing ( RNA-seq ) data from E . coli cells at different stages of growth . Inspection of the ChIP data revealed that DNA upstream of cbpA regulatory region was a target for Fis in all three ChIP studies . Strikingly , the single nucleotide resolution mapping of Fis binding performed by Kahramanoglou et al [24] identified exactly the Fis binding site defined by our in vitro footprinting analysis ( Figure S5 ) . Examination of RNA-seq data revealed that no transcripts originate from the cbpA P6 promoter during rapid growth when Fis levels are high ( Figure S5 ) . Conversely , in starved E . coli cells that contain low levels of Fis , transcription from the P6 promoter was evident ( Figure S5 ) . We reasoned that we should be able to “transplant” the repressive effect of the cbpA Fis binding region to unrelated promoters . Furthermore , we expected that the Fis binding region would only inhibit promoters active during periods of growth , when Fis levels are high . Thus , we selected four promoter regions with different regulatory properties . Three of the promoters that we selected ( galP1 , ynfE and yeaR ) are σ70 dependent promoters active in growing cells . Transcription from galP1 is stimulated by CRP , ynfE is FNR dependent and yeaR transcription requires NarL [25]–[27] . The fourth promoter that we selected ( aer ) is σ28 dependent and active during periods of starvation [28] . We compared LacZ expression from each promoter either in the presence or absence of the cbpA Fis binding element . When present the Fis binding region was placed between the promoter transcription start site and the lacZ start codon . The data show that the Fis binding element from the cbpA regulatory region repressed LacZ expression from all three “growth phase” promoters . Conversely , the aer promoter , which is active during periods of starvation , was not repressed ( Table 1 ) . We next turned our attention to the DNA binding properties of CbpA in starved E . coli . To investigate DNA binding in vivo we utilised chromatin immunoprecipitation and DNA microarrays ( ChIP-chip ) . In a preliminary analysis we compared CbpA binding in wild type E . coli BW27784 and the cbpM derivative MC108 . The data show similar patterns of CbpA binding across the E . coli chromosome in both cases ( Figure S6 ) . Recall that CbpM is known to interfere with DNA binding by CbpA . Hence , in vivo sequence requirements for CbpA-DNA interactions were determined using ChIP-chip data generated in the absence of CbpM . Because CbpA was isolated on the basis of its propensity to bind curved DNA in vitro we aligned the CbpA ChIP-chip profile with profiles of predicted DNA curvature [29] and DNA GC content . Figure 5Ai gives an overview of the alignments for the whole genome . Figure 5Aii provides a more detailed view for a smaller segment of the chromosome . Two characteristics of CbpA binding are apparent . First , CbpA binding is biased towards the Ter macrodomain ( Figure 5Ai ) . Second , there is a positive correlation between CbpA binding and DNA curvature ( Figure 5Ai and 5Aii ) . This pattern of DNA binding was also confirmed in vitro for selected targets ( Figure S7 ) . To better ascertain the relationship between DNA sequence and CbpA binding we grouped all probes on the DNA microarray according to their percentage GC content . For each group of probes we then calculated the mean CbpA binding signal . The results of this analysis confirm that CbpA binding is greatly reduced at GC-rich DNA sequences ( Figure 5B ) . Optimal CbpA binding was observed at DNA with a GC content of 43% . The average GC content of the E . coli genome is 51% . We note that , whilst CbpA seldom binds to regions that are not intrinsically curved , not all regions of predicted curvature are bound by CbpA ( Figure 5Aii ) . This is likely due to competition with other proteins that recognise curved DNA . Similarly , the in silico DNA curvature predictions may not hold true at all locations in vivo . Numerous studies have shown that entry to stationary phase results in a marked decrease in DNA supercoiling [2] , [30] , [31] . Thus , we investigated the effect of CbpA , and as a control Dps , on DNA topology in vivo using a reporter plasmid . After extraction from three day old cultures plasmid topoisomers were separated on a 1% agarose gel containing 2 . 5 µg/ml chloroquine . A sample of plasmid DNA isolated from log phase cells was also analysed as a control . As expected , the plasmid isolated from growing cells was considerably more supercoiled than the plasmid isolated from starved cells ( Figure 5Ci , compare lanes 1 and 5 ) . Remarkably , starved cells lacking dps or cbpA yielded a split distribution of plasmid topoisomers . Some plasmids were highly supercoiled with similar topology to plasmids isolated from growing cells ( compare lanes 2 , 3 and 5 ) . Other plasmids in the sample were relaxed , as seen for starved cells ( compare lanes 1–3 ) . Plasmids isolated from the ΔdpsΔcbpA strain had similar topology to the plasmids isolated from the individual gene knockout strains ( compare lanes 2–4 ) . These effects were specific to plasmids isolated from starved cells; plasmid topoisomers obtained from growing cultures of each strain did not significantly differ ( Figure 5Cii ) .
In wild type E . coli cells cbpA transcription initiates from the P1 and P2 promoters with the σ38 dependent P2 promoter being dominant [16] , [32] ( Figure 2C ) . However , during rapid growth , Fis is required to prevent uncontrolled transcription of cbpA ( Figure 1Aii ) . This uncontrolled transcription is primarily driven by the P6 promoter , which is located in the yccE gene , more than 250 base pairs upstream of cbpA . Whilst Fis prevents transcription from P6 in vivo ( Figure 2C , Figure 3B and Figure 4 ) and in vitro ( Figure 3A ) Fis appears unable to prevent RNA polymerase binding at the P6 promoter ( Figure 2B ) . This is not surprising since the DNA element bound by Fis is located ∼60 base pairs downstream of the P6 transcription start . We conclude that Fis must either act as a “road block” , to prevent transcription elongation , or prevent RNA polymerase escape from the P6 promoter . We currently favour the latter model since we were unable to detect “road blocked” transcripts in our in vitro transcription assay ( Figure 3A ) . Note that Fis also inhibits transcription from the weak P4 promoter ( Figure 2C , Figure 3A and Figure S4 ) . In this case the P4 −10 element is embedded within the primary Fis binding sequence ( Figure S4 ) . Hence , Fis binding to this site prevents RNA polymerase association with the P4 promoter ( Figure 2B lanes 5–7 ) . Our model for Fis regulation of cbpA is outlined in Figure 6A . This regulatory mechanism is salient since it is becoming increasingly apparent that bacterial chromosomes contain many intragenic promoters [33]–[35] . Thus , mechanisms must exist to ensure that these promoters do not adversely affect transcription of neighbouring operons . We speculate that Fis , and other nucleoid proteins , may serve such a purpose . Taken together with previous work our results provide a detailed molecular explanation for the phenomenon of nucleoid reorganisation that occurs in starved E . coli cells [36] . We propose that regulatory crosstalk between nucleoid proteins plays a pivotal role . In our model , Fis sits at the fulcrum of the regulatory process by binding low affinity elements overlapping the dps promoter [8] ( to inhibit σ70 dependent transcription ) and by binding high affinity sites in the cbpA regulatory region ( Figure 1 ) ( to insulate cbpA from σ70 dependent transcription ) . As cells divide , and Fis levels decrease , the different affinity of Fis for the dps and cbpA regulatory regions contributes to staged induction of dps and cbpA ( Figure 6B ) . Interestingly , despite their similar DNA binding properties , CbpA and Dps both have distinct stress response functions . Thus , CbpA can function as a co-chaperone , by virtue of its N-terminal J-domain , whilst Dps has a bacterioferritin like fold and can sequester Fe2+ ions [7] , [9] . We speculate that , to fully understand the function of CbpA and Dps in stationary phase , the relationship between their different activities will need to be unravelled .
Bacterial strains , plasmids and oligonucleotide sequences are listed in Table S1 . All cbpA regulatory region sequences are numbered with respect to the P1 transcription start point ( +1 ) and with upstream and downstream locations denoted by ‘−’ and ‘+’ prefixes respectively . Activities are shown in Miller units [37] and are the average of three or more independent experiments with a standard deviation of <10% . Background LacZ activity values , generated from cells carrying a “promoterless” pRW50 , were subtracted . Cells were grown aerobically , at 37°C , in LB media . Assays were performed using either JCB387 or the derivative JCB3871Δfis . Transcript start sites were mapped by primer extension , as described in Lloyd et al . [38] , using RNA purified from strains carrying the 302 base pair cbpA regulatory DNA fragment cloned in pRW50 . The 5′ end-labelled primer D49724 , which anneals downstream of the HindIII site in pRW50 was used in all experiments . Primer extension products were analysed on denaturing 6% polyacrylamide gels , calibrated with arbitrary sequencing reactions , and visualized using a Fuji phosphor screen and Bio-Rad Molecular Imager FX . CbpA and derivatives were all purified as described [15] . Fis and RNA polymerase were prepared as described previously [8] . EMSA , DNAseI and KMnO4 footprinting with Fis and/or RNA polymerase are described by Grainger et al . [8] . The in vitro transcription experiments were performed as described [8] using the system of Kolb et al . [39] . Protein and DNA concentrations used for all in vitro experiments are provided in the figure legends . Chromatin Immunoprecipitation was done exactly as described previously [40] . Formaldehyde crosslinked nucleoprotein obtained from stationary phase BW27784 or MC108 cells was fragmented by sonication and CbpA-DNA complexes were precipitated using a rabbit polyclonal antibody against CbpA . A control mock immunoprecipitation ( from which anti-CbpA was omitted ) was done in parallel . The “plus and minus antibody” DNA samples were then labelled with Cy5 and Cy3 respectively before being mixed and hybridised to a 43 , 450 feature DNA microarray ( Oxford Gene Technology ) . After hybridisation , washing and scanning the Cy5 and Cy3 signal was calculated for each probe on the array ( Table S2 ) . The MC108 experiment was done in duplicate , and an average Cy5/Cy3 ratio was used for further analysis ( referred to as the CbpA binding signal ) . The images shown in Figure 4 were generated using DNA plotter software [41] . To facilitate detailed inspection of the CbpA ChIP-chip data a file that can be loaded into DNA plotter or the Artemis genome browser is provided in the supplementary material ( Table S3 ) . These data should be loaded into the software as a graph after first installing the E . coli K-12 genome sequence ( provided as a genbank file in Table S4 ) . The DNA curvature analysis of the E . coli chromosome was done using the CURVATURE software package [42] exactly as described previously [29] . The DNA GC profile was calculated using the internal graph function in DNA plotter . We note that low GC content is not an absolute indicator of increased DNA curvature on a local scale of a few base pairs . However , for the large segments of DNA considered here , there is a clear inverse correlation between GC content and DNA curvature . We monitored superhelicity of plasmid pJ204 in strain BW27784 and the Δdps , ΔcbpA or ΔdpsΔcbpA derivatives . Transformants were grown in LB medium at 37°C for three days . Plasmid DNA samples were prepared using a QIAprep Spin Miniprep Kit ( Qiagen ) . Topoisomers were separated on a 1% agarose gel containing 2 . 5 µg/ml chloroquine . Gels were run for 60 hours at 40 V in the dark . After washing with water for at least 2 hours the gel was stained with ethidium bromide for 2 hours and photographed under UV illumination . Overnight cultures of E . coli BW25113 were diluted 1∶100 into fresh LB media and grown at 37°C with aeration . Samples ( 1 ml ) were taken at the indicated OD650 values and the cells harvested by centrifugation . Cells were re-suspended in Laemmli buffer so that the number of cells in each sample was equivalent . After boiling for 10 min cytoplasmic proteins were separated by SDS-PAGE and Fis was detected using Western immunoblotting as previously described [43] . The blots were also probed using antibodies against RpoA ( Neoclone ) as a control . | Compaction of chromosomal DNA is a fundamental process that impacts on all aspects of cellular biology . However , our understanding of chromosome organisation in bacteria is poorly developed . Since bacteria are amongst the most abundant living organisms on the planet , this represents a startling gap in our knowledge . Despite our lack of understanding , it has long been known that Escherichia coli , and other bacteria , radically re-model their chromosomes in response to environmental stress . This is most notable during periods of starvation , when the E . coli chromosome is super compacted . In dissecting the molecular mechanisms that control this phenomenon , we have found that regulatory cross-talk between DNA–organising proteins plays an essential role . Thus , the major DNA folding protein from growing E . coli inhibits production of the major chromosome organisers in starved cells . Our findings illustrate the highly dynamic nature of bacterial chromosomes . Thus , DNA topology , gene transcription , and chromosome folding proteins entwine to create a web of interactions that define the properties of the chromosome . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"biochemistry",
"model",
"organisms",
"genetics",
"biology",
"genomics",
"microbiology",
"genetics",
"and",
"genomics"
]
| 2013 | E. coli Fis Protein Insulates the cbpA Gene from Uncontrolled Transcription |
Extracellular cues affect signaling , metabolic , and regulatory processes to elicit cellular responses . Although intracellular signaling , metabolic , and regulatory networks are highly integrated , previous analyses have largely focused on independent processes ( e . g . , metabolism ) without considering the interplay that exists among them . However , there is evidence that many diseases arise from multifunctional components with roles throughout signaling , metabolic , and regulatory networks . Therefore , in this study , we propose a flux balance analysis ( FBA ) –based strategy , referred to as integrated dynamic FBA ( idFBA ) , that dynamically simulates cellular phenotypes arising from integrated networks . The idFBA framework requires an integrated stoichiometric reconstruction of signaling , metabolic , and regulatory processes . It assumes quasi-steady-state conditions for “fast” reactions and incorporates “slow” reactions into the stoichiometric formalism in a time-delayed manner . To assess the efficacy of idFBA , we developed a prototypic integrated system comprising signaling , metabolic , and regulatory processes with network features characteristic of actual systems and incorporating kinetic parameters based on typical time scales observed in literature . idFBA was applied to the prototypic system , which was evaluated for different environments and gene regulatory rules . In addition , we applied the idFBA framework in a similar manner to a representative module of the single-cell eukaryotic organism Saccharomyces cerevisiae . Ultimately , idFBA facilitated quantitative , dynamic analysis of systemic effects of extracellular cues on cellular phenotypes and generated comparable time-course predictions when contrasted with an equivalent kinetic model . Since idFBA solves a linear programming problem and does not require an exhaustive list of detailed kinetic parameters , it may be efficiently scaled to integrated intracellular systems that incorporate signaling , metabolic , and regulatory processes at the genome scale , such as the S . cerevisiae system presented here .
Intracellular biochemical networks are comprised of signaling , metabolic , and regulatory processes . ( Note that here we use “regulation” to refer specifically to transcriptional regulatory and protein synthesis networks , and “signaling” to describe intracellular reactions that drive responses to the extracellular environment . ) Until recently , computational analyses focused independently on signaling , metabolic , and regulatory networks . However , high-throughput experimental data coupled with computational systems analysis techniques have elucidated multifunctional components involved in fundamental disease processes [1]–[4] . For example , signaling cascades are triggered by the presence of extracellular stimuli and often result in activation of transcription factors . These transcription factors function in regulatory networks , regulating the transcription of associated genes and the synthesis of various proteins used in signal transduction and metabolism . Cellular metabolism is responsible for the production of energy in the form of adenosine triphosphate ( ATP ) and the synthesis of amino acids among other biomass precursors , all of which are used elsewhere in the cell . Consequently , a key challenge in the post-genomic era is to consider the interconnectedness of biochemical networks and how extracellular cues affect highly integrated intracellular processes to elicit cellular responses such as growth or differentiation . Dynamic [5] , [6] and structural analyses [7] have been employed to quantitatively analyze large-scale biochemical network modules . Typically , in dynamic analyses , a set of ordinary differential equations ( ODEs ) describing the mass ( balance ) of each species in the system is constructed . Despite its generality , the application of this type of mechanistic model at a genome-scale is largely considered impractical because it necessitates the consideration of many pathways for which detailed reactions and their kinetic parameters are not yet known . Structural analyses like flux balance analysis ( FBA ) can calculate phenotypic properties of a biological network like a steady-state flux ( i . e . , reaction rate ) distribution without detailed kinetic information . FBA requires a physiologically relevant objective function ( e . g . , in the case of metabolism , maximizing the growth rate or maximizing ATP production ) , mass-balance constraints ( i . e . , the stoichiometry of the reactions ) , and constraints on reaction directions and thermodynamics . Since the physicochemical constraints are readily defined ( e . g . , from the annotated genome sequence and measured enzymatic capacities ) , FBA has been used effectively to study large-scale biochemical networks , particularly metabolic networks [8] . However , in general , the steady-state assumption of FBA prevents it from generating dynamic concentration profiles of intracellular species . An additional challenge to the modeling of integrated systems is that time scales of intracellular biochemical networks generally span multiple orders of magnitude . Signaling and metabolic reactions typically occur rapidly . For example , kinase/phosphatase reactions , protein conformational changes , and most metabolic reactions occur on the order of fractions of a second to seconds [9] . By contrast , receptor internalization [10] and regulatory events [11] , [12] , as well as end-stage phenotypic properties such as cellular growth or differentiation [13] can take several minutes to hours . These multiple time-scales pose computational challenges for the quantitative analysis of integrated systems . For instance , kinetic model-based strategies suffer from a scarcity of values for kinetic parameters as well as poor accuracy of known kinetic parameters [14] . In addition , models of integrated systems are inherently “stiff , ” i . e . , they must include “fast” and “slow” reaction dynamics simultaneously [15] , and they are consequently difficult to simulate and extremely sensitive to modeling errors [16] . Indeed , it is challenging to apply FBA to models of integrated systems because of the steady-state assumption intrinsic to FBA and the “fast” and “slow” reaction dynamics that coexist intracellularly . Due to these complexities , previous models and analyses have focused primarily on network modules rather than integrated systems . These include kinetic , stoichiometric , and causality analyses of modular signaling systems [17]–[20] , metabolism [21]–[25] , and regulation [26] , [27] . Some preliminary dynamic analyses of integrated systems have been completed . Integrated analyses of regulatory and metabolic networks revealed novel mechanisms in Saccharomyces cerevisiae and Escherichia coli [28]–[30] . Metabolic reactions were represented stoichiometrically , and regulatory reactions were captured by representing gene regulatory rules using a Boolean formalism . FBA was implemented assuming quasi-steady-state conditions , i . e . , the typical time constant of metabolic transients was relatively faster than the simulation time step for temporal integration of phenotypic variables ( e . g . , biomass as a measure of cellular growth ) . Recently , a kinetic model accounting for signal transduction , metabolism , and regulation was constructed to describe the response of S . cerevisiae to osmotic shock [31] . This model connected specific outputs of one network ( e . g . , signaling ) with the inputs of another network ( e . g . , metabolism ) in a “sequential” fashion . The complete set of interactions among the biochemical networks , such as feedback and feed-forward of proteins expressed as a function of the regulatory network to signaling and metabolism , was not considered . Additionally , it required an exhaustive list of kinetic parameters ( e . g . , rate constants ) for the reactions , and time-courses of individual modules ( or collections of reactions ) were evaluated separately . As reconstructions of large-scale signaling and metabolic networks are emerging , there is a growing need for the development of a framework to study these networks from an integrated perspective [9] . The purpose of this study was to develop a FBA-based computational framework , termed integrated dynamic Flux Balance Analysis ( idFBA ) , for the quantitative , dynamic analysis of cellular behaviors arising from signaling , metabolic , and regulatory networks at the genome-scale . The idFBA framework requires an integrated stoichiometric reconstruction of signaling , metabolic , and regulatory processes . It assumes quasi-steady-state conditions for “fast” reactions and incorporates “slow” reactions in a time-delayed manner . To assess the efficacy of idFBA , we developed a prototypic integrated system with topological features characteristic of those observed in existing signaling , metabolic , and regulatory network reconstructions as well as kinetic parameters reported in literature . Additionally , we applied in a similar manner the idFBA framework to a representative module in S . cerevisiae as a validation of our approach . idFBA allowed for quantitative , dynamic analysis of systemic effects of extracellular cues on phenotypes of these systems and generated comparable time-course predictions when contrasted with kinetic models . Ultimately , we demonstrate how idFBA enables genome-scale quantitative , dynamic analysis of integrated systems .
In order to assess the efficacy of idFBA , a prototypic integrated system was constructed with characteristics typical of those observed in published reconstructions of signaling , metabolic , and regulatory networks ( see Figures 1 and 2 ) . Specifically , we generated representative reactions with stoichiometric relationships and estimated their associated rate constants from literature . Here we briefly describe the reactions that are considered in each network and their typical time scales . Detailed information on these reactions and associated kinetic parameters is provided in Text S1 . To assess the applicability of idFBA to actual biological systems , a representative integrated module in S . cerevisiae , the prototypic single-cell eukaryote , was investigated . This module was comprised of key aspects of yeast osmoregulation , i . e . , the active processes with which yeast cells monitor and adjust pressure and control their shape , turgor , and water content in response to extracellular conditions [34] . The signaling , metabolic , and regulatory activities included in this module are illustrated in Figure 3 . Specifically , we reconstructed a portion of the high-osmolarity glycerol response ( HOG ) pathway , one of four major mitogen-activated protein ( MAP ) kinase cascades in S . cerevisiae , from existing literature . The HOG MAP kinase pathway plays a pivotal role in the adaptation of S . cerevisiae to conditions of high external osmolarity [35] . For example , yeast cells deficient in this pathway cannot proliferate on media containing high levels of osmotically active molecules [34]–[36] . Extensive genetic analysis has previously been performed , leading to experimental identification of many activating and inhibiting components of the HOG signaling pathway [37] . In general , yeast cells use the HOG pathway to accumulate glycerol under hyperosmotic conditions , to balance the osmotic pressure with the extracellular environment . Osmotic stress signals are communicated via the HOG signaling pathway , leading to the activation of Hot1 and other transcription factors . These transcription factors subsequently promote the expression of glycolytic enzymes , such as Stl1 , Gpd1 , and Gpp2 , thereby catalyzing metabolic reactions leading to increased glycerol production . As this model serves an illustrative purpose here , the HOG pathway was restricted to the key set of reactions necessary for its phenotypic function . Specifically , 26 reactions spanning 48 components were assimilated in stoichiometric matrix form , including 16 reactions across 33 components in signaling; a single transcription factor activating three regulated genes; and seven reactions across 12 components in metabolism . Inputs of this module included osmotic shock ( signaling ) and glucose ( metabolism ) , and outputs included glycerol ( metabolism ) . Key reactions connecting the underlying signaling , metabolic , and regulatory processes were the translocation of the kinase Hog1 into the nucleus for the activation of transcription factor Hot1 ( signal transduction and metabolism ) , and the synthesis of metabolic enzymes Stl1 , Gpd1 , and Gpp2 for reactions involved in the conversion of glucose to glycerol ( transcriptional regulation and metabolism ) . Other reactions in the HOG pathway as previously experimentally characterized ( e . g . , inhibition of Hog1 by phosphatases Ptp2 , Ptp3 , and Ptc1 , thereby allowing the cell to keep the HOG pathway in check and maintain osmotic balance ) were excluded from the reconstruction used here for simplicity . As with the prototypic system , the representative integrated yeast module was implemented using the idFBA framework as well as a kinetic model similar to the one in [31] , and the two approaches were contrasted for validation purposes . Rate constants describing the kinetics of the system were culled from available experimental data , notably [31] . For complete details of the reconstructed yeast HOG pathway , including listings of reactions , rate constants , and kinetic equations , see Text S2 . One modeling technique for evaluating cellular phenotypes is called flux balance analysis ( FBA ) . FBA is a constraints-based approach that attempts to derive a phenotype in the form of a steady-state flux distribution for the reactions in a given biological system . FBA is based on the principle that all expressed phenotypes of a given biological system must satisfy basic constraints that are imposed on the functions of all cells [8] , [38] , [39] . These constraints are physico-chemical ( i . e . , physical laws like conservation of mass and energy ) ; topological ( i . e . , spatial restrictions on metabolites within cellular compartments ) ; and environmental ( i . e . , nutrient availability , pH , and temperature , all of which vary over time and space ) [8] , [20] , [38] . Because FBA yields fluxes rather than concentrations , limited kinetic information is required for its implementation . FBA requires a stoichiometric reconstruction of the biochemical network of interest . An annotated genome cataloging which reactions specific enzymes catalyze is the basis for a detailed description of a network's components and interactions [40] . This biochemical network reconstruction can be represented in matrix form , S , where S is of size m components×n reactions and is comprised of stoichiometric coefficients that capture the underlying reactions of the biochemical network . After the network is reconstructed , fluxes are calculated by deriving a dynamic mass balance for all the components within the system [7] , [38] . Specifically , at steady state , the change in the amount of a component C over time t across all reactions within the system must be zero . Consequently , mass balance is defined in terms of the flux through each reaction and the stoichiometry of that reaction , and a set of coupled ordinary differential equations relating the roles of reactions with components may be written in the form of Equation 3 . ( 3 ) Here , S denotes the m×n matrix of stoichiometric coefficients and v denotes the vector of n reaction fluxes , with each element ( row ) of the n-row vector v corresponding to the flux in the associated reaction ( column ) in S . The vector C is a m-row vector defining the concentrations of the m components within the system . This mass balance represents the principal constraint in FBA and defines a feasible solution space for the set of fluxes . Additional constraints such as thermodynamics can be incorporated into FBA as well , further narrowing the possible distribution of fluxes [24] , [41] . Equation 3 generally leads to an under-determined system because the number of components tends to be far fewer than the number of reactions . Even with additional constraints , FBA usually requires performing an optimization with linear programming ( LP ) to identify a particular flux distribution . In other words , FBA involves optimizing the set of fluxes such that the flux through a particular cellular reaction is maximized ( or minimized ) . A cellular objective represents what a given biological system has optimized toward through evolutionary pressures [42] . It is defined as a linear equation ( Equation 4 ) , where c is the vector that defines the coefficients , or weights , for each of the fluxes in v [41] . ( 4 ) This general representation of Z , wherein the elements of c can be easily manipulated , enables the formulation of many diverse objectives . Common choices for cellular objective functions in models of metabolic networks include biomass production [24] , [43] , energy [44] , and byproduct production [45] . Ultimately , FBA attempts to solve the LP problem in Equation 5 to find a physiologically-relevant cellular phenotype in the form of a flux distribution v that optimizes Z while lying in the bounded solution space defined by a set of physio-chemical , topological , and environmental constraints . ( 5 ) Note that vlb and vub are the lower and upper bounds on the reaction fluxes , respectively . For example , thermodynamic constraints or reaction directionalities can be incorporated by setting a given vlb = 0 . Though the steady-state assumption of FBA precludes the calculation of dynamic concentrations of the network components , dynamic profiles of cellular phenotypes ( e . g . , cellular growth or differentiation ) have been successfully predicted with a quasi-steady-state assumption [24] , [25] , [29] . This assumption involves discretizing the time domain into intervals , and ( 1 ) solving the LP problem contained within FBA at the beginning of each interval , and ( 2 ) based on the resultant flux data , solving a system of ODEs for concentrations over time within each interval . Applications of FBA to dynamic simulations have focused on metabolic networks because time constants of metabolic transients are typically very rapid when contrasted with time constants characterizing whole-cell phenotypic changes . Exceptions include the incorporation of gene regulatory events , which are much slower than metabolic reactions , into FBA for time-course simulation of metabolic reactions [28] , [29] . In these cases , the regulatory constraints were described as Boolean operators and imposed in a time-delayed manner . However , these examples are limited to metabolic and regulatory processes and do not consider changes in the mass balance ( e . g . , protein synthesis ) arising from the interactions between metabolic and regulatory processes and signaling systems . Consequently , quantitative , dynamic analyses of integrated cellular systems have not been explored in detail , limiting the characterization of whole-cell function . As previously described , the stoichiometric reconstruction enforces explicit , chemically-consistent accounting of the components and reactions underlying a biochemical network , and facilitates the systematic analysis of fundamental network properties with FBA and associated analysis techniques [32] . The stoichiometric reconstruction and FBA are particularly applicable to large-scale networks , for which a lack of kinetic data ( e . g . , rate constants ) makes kinetic-based approaches impractical . Indeed , stoichiometric reconstruction and FBA have been applied successfully to large-scale metabolic and signaling networks , elucidating characteristics of these networks [8] , [9] . Therefore , integrating signaling reactions with metabolic and regulatory reactions using FBA can facilitate the dynamic analysis of cellular phenotypes arising from environmental cues and provide a complete snapshot of cellular sysems . However , as previously described , applying FBA directly to integrated networks is challenging . First , unlike metabolic systems in which objectives for the FBA formulation are often experimentally characterized ( e . g . , the production of biomass ) , objectives of signaling and regulatory systems are not well-defined . Second , integrated networks are comprised of reactions with mixed time scales ( e . g . , signaling reactions are generally much faster than regulatory reactions ) , and FBA has previously been applied only to fast reactions for which steady-state assumptions hold . Here we describe the idFBA framework , including how we address these challenges . We use the prototypic integrated system as the basis for this discussion . To validate the results of the idFBA framework , we developed kinetic models of the prototypic integrated system and the representative integrated yeast module . As previously stated , kinetic models describe the temporal changes of compound concentrations due to production , degradation , modification , or transport [46] . In other words , the rate of change of the concentration Ci of the ith compound within a system may be described as in Equation 11 below [46] . Here Sij is the stoichiometric coefficient , vj is the rate of the jth reaction , and n is the total number of reactions in the network . Reactions that produce or consume the ith compound have non-zero stoichiometric coefficients and are therefore included in the ith differential equation . ( 11 ) The reaction rates for the network , v , are functions of component concentrations , such as the concentrations of enzymes ( e . g . , kinases and phosphatases within a signaling network ) , as well as parameters including kinetic constants . These rates are described by different types of kinetic laws . For example , Michaelis-Menten expressions can be used to model enzyme kinetics [47] , [48] . Our ODE models of the prototypic integrated system and representative integrated yeast module were constructed from the underlying reaction network , with rate constants ( i . e . , kinetic parameters ) obtained from literature . The systems of ODEs were continuously solved over the time window of interest ( equivalent to that of the corresponding idFBA implementations ) . Details of these models , including the kinetic equations , kinetic constants , and ordinary differential equations , are presented in Text S1 and S2 . It is important to note that idFBA and kinetic modeling constitute two independent approaches . The idFBA framework involves performing an optimization , over multiple discretized time steps , to approximate the dynamics of a system with time-delay information from strictly stoichiometric constraints . By contrast , a kinetic model requires all of the kinetic parameters and , by continuously solving a set of ordinary differential equations , yields a more detailed portrait of the system dynamics . We attempt to illustrate here how the idFBA framework , with significantly fewer parameters , approximates the system dynamics observed through much more detailed ODE models . Technical implementation details . The kinetic models of the integrated prototypic system and representative integrated yeast module were implemented using the ode23tb ODE solver in MATLAB v . 7 . 5 ( part of the MathWorks R2007b release package ) . The ode23tb solver is an implementation of an implicit Runge-Kutta formula , comprised of a trapezoidal rule followed by a backward differentiation formula of order two . The solver compromises efficiency for crude tolerances [49] .
The integrated dynamic Flux Balance Analysis ( idFBA ) framework presented here couples stoichiometric reconstructions of signaling , metabolic , and transcriptional regulatory networks with Flux Balance Analysis ( FBA ) to predict dynamic profiles of cellular phenotypes as a function of extracellular stimuli . Instantaneous inclusion of “slow” reactions in a time-delayed fashion accounted for network interactions occurring over a wide range of time scales . Previous approaches based on FBA have only addressed the coupling of regulatory structure with metabolic systems [29] , which do not account for the effects of extracellular signaling cues on cellular phenotype . The key features and results described here include: ( 1 ) an explicit accounting of the protein synthesis demands of a transcriptional regulatory network in the context of signaling and metabolic functions; ( 2 ) a quasi-steady-state description of cellular signaling events , readily interfaced with metabolic and regulatory networks; ( 3 ) similar dynamic profiles of phenotypic variables ( e . g . , biomass production ) between the idFBA framework presented here and an explicit kinetic model; and ( 4 ) applicability of the idFBA framework to actual biological systems through an illustrative example using yeast osmoregulation and agreement with published values . To implement idFBA , the objective function for the underlying optimization problem included , for signaling networks , the reactions associated with the activation of transcription factors . The subsequent analysis resulted in “excluded reaction fluxes” ( e . g . , receptor internalization and protein degradation ) . These reactions were specified as “active” to denote their participation in the reaction network by imposing simple constraints ( v = 0 ) on their counterparts , as described in “Conceptual Methods and Framework . ” Comparison with the detailed kinetic model validates the idFBA approach . Specifically , approximating the temporal progression of “slow” reactions in signaling , metabolic , and regulatory networks as steady-state constraints with time-delay and duration parameters provides acceptable predictions of the dynamic trends of a cell's phenotypic behavior . The primary motivation for comparing idFBA with a detailed kinetic model was to determine whether idFBA would yield comparable temporal behavior in spite of the inherent approximation it contained . Optimization-based approaches have provided accurate quantitative predictions of cellular growth [24] . However , signaling networks have not previously been modeled at a scale comparable to that of metabolic and regulatory networks [9] . Databases are increasingly available for signaling networks and efforts are ongoing to reconstruct larger , genome-scale signaling systems [32] . As this information becomes available , the idFBA framework can be applied to cellular systems and be coupled with experimental assays to generate quantitative hypotheses and assist in an iterative model-building process for deriving emergent properties of these systems . The idFBA framework optimizes the system at the current time step , tcurrent , according to the linear programming formulation of FBA ( Equation 17 ) . ( 17 ) Altering this framework to impose a multi-horizon formulation may facilitate the evaluation of different objective functions because the formulation naturally accounts for long-term effects of the calculated flux distribution at the current time step [25] . The multi-horizon formulation is shown in Equation 18 , where wj is the weight associated with the objective after Tj sample times . ( 18 ) The main assumption of a multi-horizon formulation is that the flux distribution at tcurrent is determined such that it maximizes a cellular objective within a certain future time period of interest . The resulting optimization problem , including a Boolean representation of the transcriptional regulatory network , becomes a multi-horizon , mixed-integer linear programming problem . Though the solution of such a problem needs further development for its scalability to large-scale systems [64] , it may shed light on whether signal transduction at the current time step is optimally driven by a long-term objective . Currently , multi-stage optimization problems have been solved only for metabolic systems [25] . Recently , a method called Biological Objective Solution Search ( BOSS ) was developed for the inference of an objective function for a biological system from its underlying network stoichiometry as well as experimentally-measured flux distributions [54] . This method identifies objectives from experimental fluxes by defining a putative stoichiometric objective reaction , adding this reaction to the existing set of stoichiometric constraints , and maximizing it via linear programming . This new approach is capable of inferring the objective functions of metabolic networks , as well as metabolic and regulatory networks for which the objective is not well-characterized experimentally . Therefore , utilizing BOSS to identify objectives for the signaling , regulatory , and metabolic networks would facilitate the identification of an in silico flux distribution for the integrated system , a key step in the idFBA framework . The fact that different reactions occur on different time scales ( e . g . , signaling reactions are usually fast whereas regulatory reactions are usually slow ) is readily handled within the idFBA framework . Reactions with time constants of more than a unit time step are considered “slow” . However , identifying the optimal discretization of the time domain would facilitate a more accurate simulation for systems with multiple time-scales . Given typical rate constants , model reduction [15] and Monte Carlo sampling [65] techniques may help characterize representative time-scales of a given system as well . As illustrated by the idFBA results for the prototypic integrated system and particularly the representative yeast module , the methodology and analyses afforded by this framework can provide insight into fundamental characteristics of biological systems , including network components and interactions . Evaluating how whole-cell systems respond to different perturbations , including modifications to environmental cues as well as intracellular reactions , can offer insights into disease mechanisms and possible therapeutic avenues . For example , assessing how genetic perturbations of signaling proteins affect the transcriptional program and metabolism of a cell is essential to fully appreciating the end-stage phenotypic effects of the perturbations on the whole cell . Furthermore , evaluating how modifications to an existing transcriptional regulatory program ( e . g . , altering the Boolean rules governing transcription of one or more genes ) affect whole-cell behavior is essential in the design and engineering of metabolic systems . Such a complete picture of cellular response can drive accurate predictions of disease and drug discovery . Additionally , unlike kinetic-based models and other similar approaches , the idFBA framework requires significantly fewer parameters and can facilitate an approximation of the dynamics of large-scale systems quickly and efficiently , given a stoichiometric network reconstruction . As has been hypothesized in the literature recently , our idFBA results support the theory that the structure of a network , rather than the detailed kinetic values that describe it , can drive the dynamics of its phenotype [66] . In conclusion , a novel technique called integrated dynamic Flux Balance Analysis ( idFBA ) has been developed to analyze integrated systems , and specifically to account for the interactions between signaling , metabolic , and transcriptional regulatory networks across many time scales . This approach facilitates the study of systemic effects of extracellular cues on cellular behavior in a quantitative manner . Additionally , the success of idFBA on a prototypic integrated system as well as a representative integrated yeast module serves as a benchmark for future analyses of integrated biochemical systems . | Cellular systems comprise many diverse components and component interactions spanning signal transduction , transcriptional regulation , and metabolism . Although signaling , metabolic , and regulatory activities are often investigated independently of one another , there is growing evidence that considerable interplay occurs among them , and that the malfunctioning of this interplay is associated with disease . The computational analysis of integrated networks has been challenging because of the varying time scales involved as well as the sheer magnitude of such systems ( e . g . , the numbers of rate constants involved ) . To this end , we developed a novel computational framework called integrated dynamic flux balance analysis ( idFBA ) that generates quantitative , dynamic predictions of species concentrations spanning signaling , regulatory , and metabolic processes . idFBA extends an existing approach called flux balance analysis ( FBA ) in that it couples “fast” and “slow” reactions , thereby facilitating the study of whole-cell phenotypes and not just sub-cellular network properties . We applied this framework to a prototypic integrated system derived from literature as well as a representative integrated yeast module ( the high-osmolarity glycerol [HOG] pathway ) and generated time-course predictions that matched with available experimental data . By extending this framework to larger-scale systems , phenotypic profiles of whole-cell systems could be attained expeditiously . | [
"Abstract",
"Introduction",
"Methods",
"Discussion"
]
| [
"cell",
"biology/cell",
"signaling",
"computational",
"biology/transcriptional",
"regulation",
"mathematics",
"computational",
"biology/metabolic",
"networks",
"computational",
"biology/signaling",
"networks",
"computational",
"biology/molecular",
"dynamics",
"computational",
"biology/systems",
"biology"
]
| 2008 | Dynamic Analysis of Integrated Signaling, Metabolic, and Regulatory Networks |
Traditional approaches to the problem of parameter estimation in biophysical models of neurons and neural networks usually adopt a global search algorithm ( for example , an evolutionary algorithm ) , often in combination with a local search method ( such as gradient descent ) in order to minimize the value of a cost function , which measures the discrepancy between various features of the available experimental data and model output . In this study , we approach the problem of parameter estimation in conductance-based models of single neurons from a different perspective . By adopting a hidden-dynamical-systems formalism , we expressed parameter estimation as an inference problem in these systems , which can then be tackled using a range of well-established statistical inference methods . The particular method we used was Kitagawa's self-organizing state-space model , which was applied on a number of Hodgkin-Huxley-type models using simulated or actual electrophysiological data . We showed that the algorithm can be used to estimate a large number of parameters , including maximal conductances , reversal potentials , kinetics of ionic currents , measurement and intrinsic noise , based on low-dimensional experimental data and sufficiently informative priors in the form of pre-defined constraints imposed on model parameters . The algorithm remained operational even when very noisy experimental data were used . Importantly , by combining the self-organizing state-space model with an adaptive sampling algorithm akin to the Covariance Matrix Adaptation Evolution Strategy , we achieved a significant reduction in the variance of parameter estimates . The algorithm did not require the explicit formulation of a cost function and it was straightforward to apply on compartmental models and multiple data sets . Overall , the proposed methodology is particularly suitable for resolving high-dimensional inference problems based on noisy electrophysiological data and , therefore , a potentially useful tool in the construction of biophysical neuron models .
Among several tools at the disposal of neuroscientists today , data-driven computational models have come to hold an eminent position for studying the electrical activity of single neurons and the significance of this activity for the operation of neural circuits [1]–[4] . Typically , these models depend on a large number of parameters , such as the maximal conductances and kinetics of gated ion channels . Estimating appropriate values for these parameters based on the available experimental data is an issue of central importance and , at the same time , the most laborious task in single-neuron and circuit modeling . Ideally , all unknown parameters in a model should be determined directly from experimental data analysis . For example , based on a set of voltage-clamp recordings , the type , kinetics and maximal conductances of the voltage-gated ionic currents flowing through the cell membrane could be determined [5] and , then , combined in a conductance-based model , which replicates the activity of the biological neuron of interest under current-clamp conditions with sufficient accuracy . Unfortunately , this is not always possible , especially for complex compartmental models , which contain a large number of ionic currents . A first problem arises from the fact that not all parameters can be estimated within an acceptable error margin , especially for small currents and large levels of noise . A second problem arises from the practice of estimating different sets of parameters based on data collected from different neurons of a particular type , instead of estimating all unknown parameters using data collected from a single neuron . Different neurons of the same type may have quite different compositions of ionic currents [6]–[9] ( but , see also [10] ) . This implies that combining ionic currents measured from different neurons in the same model or even using the average of several parameters calculated over a population of neurons of the same type will not necessarily result in a model that expresses the experimentally recorded patterns of electrical activity under current-clamp conditions . Usually , only some parameters are well characterized , while others are difficult or impossible to measure directly . Thus , most modeling studies rely on a mixture of experimentally determined parameters and estimates of the remaining unknown ones using automated optimization methodology ( see , for example , [11]–[22] ) . Typically , these methods require the construction of a cost function ( for measuring the discrepancy between various features of the experimental data and the output of the model ) and an automated parameter selection method , which iteratively generates new sets of parameters , such that the value of the cost function progressively decreases during the course of the simulation ( see [23] for a review ) . Popular choices of such methods are evolutionary algorithms , simulated annealing and gradient descent methods . Often , a global search method ( i . e . an evolutionary algorithm ) is combined with local search ( gradient descent ) for locating multiple minima of the cost function with high precision . Since a poorly designed cost function ( for example , one that merely matches model and experimental membrane potential trajectories ) can seriously impede optimization , the construction of this function often requires particular attention ( see , for example , [24] ) . Nevertheless , these computationally intensive methodologies have gained much popularity , particularly due to the availability of powerful personal computers at consumer-level prices and the development of specialized optimization software ( e . g . [25] ) . Alternative approaches also exist as , for example , methods based on the concept of synchronization between model dynamics and experimental data [26] . An emerging trend in parameter estimation methodologies for models in Computational Biology is to recast parameter estimation as an inference problem in hidden dynamical systems and then adopt standard Computational Statistics techniques to resolve it [27] , [28] . For example , a particular study following this approach makes use of Sequential Monte Carlo methods ( particle filters ) embedded in an Expectation Maximization ( EM ) framework [28] . Given a set of electrophysiological recordings and a set of dynamic equations that govern the evolution of the hidden states , at each iteration of the algorithm the expected joint log-likelihood of the hidden states and the data is approximated using particle filters ( Expectation Step ) . At a second stage during each iteration ( Maximization Step ) , the log-likelihood is locally maximized with respect to the unknown parameters . The advantage of these methods , beyond the fact that they recast the estimation problem in a well-established statistical framework , is that they can handle various types of noisy biophysical data made available by recent advances in voltage and calcium imaging techniques . Inspired by this emerging approach , we present a method for estimating a large number of parameters in Hodgkin-Huxley-type models of single neurons . The method is a version of Kitagawa's self-organizing state-space model [29] combined with an adaptive algorithm for selecting new sets of model parameters . The adaptive algorithm we have used is akin to the Covariance Matrix Adaption ( CMA ) Evolution Strategy [30] , but other methods ( e . g . Differential Evolution as described in [31] ) may be used instead . We demonstrate the applicability of the algorithm on a range of models using simulated or actual electrophysiological data . We show that the algorithm can be used successfully with very noisy data and it is straightforward to apply on compartmental models and multiple datasets . An interesting result from this study is that by using the self-organizing state-space model in combination with a CMA-like algorithm , we managed to achieve a dramatic reduction in the variance of the inferred parameter values . Our main conclusion is that a large number of parameters in a conductance-based model of a neuron ( including maximal conductances , reversal potentials and kinetics of gated ionic currents ) can be inferred from low-dimensional experimental data ( typically , a single or a few recordings of membrane potential activity ) using the algorithm , if sufficiently informative priors are available , for example in the form of well-defined ranges of valid parameter values .
We begin by presenting the current conservation equation that describes the time evolution of the membrane potential for a single-compartment model neuron: ( 1 ) where , and are all functions of time . In the above equation , is the membrane capacitance , is the membrane potential , is the externally applied ( injected ) current , and are the maximal conductance and reversal potential of the leakage current , respectively , and is the transmembrane ionic current . A voltage-gated current can be modeled according to the Hodgkin-Huxley formalism , as follows: ( 2 ) where and are both functions of time . In the above expression , and are the maximal conductance and reversal potential of the ionic current , and are dynamic gating variables , which model the voltage-dependent activation and inactivation of the current , and is a small positive integer power ( usually , not taking values larger than 4 ) . The product is the proportion of open channels in the membrane that carry the current . The gating variables and obey first-order relaxation kinetics , as shown below: ( 3 ) where the steady states ( , ) and relaxation times ( , ) are all functions of voltage . Using vector notation , we can write the above system of Ordinary Differential Equations ( ODEs ) in more concise form: ( 4 ) where the state vector is composed of the time-evolving state variables , and and the vector-valued function , which describes the evolution of in time , is formed by the right-hand sides of Eqs . 1 and 3 . Notice that also depends on a parameter vector , which for now is dropped from Eq . 4 for notational clarity . Components of are the maximal conductances , the reversal potentials and the various parameters that control the voltage-dependence of the steady states and relaxation times in Eq . 3 . The above deterministic model does not capture the inherent variability in the electrical activity of neurons , but rather some average behavior of intrinsically stochastic events . In general , this variability originates from various sources , such as the random opening and shutting of transmembrane ion channels or the random bombardment of the neuron with external ( e . g . synaptic ) stimuli [32] . Here , we model the inherent variability in single-neuron activity by augmenting Eq . 4 with a noisy term and re-writing as follows: ( 5 ) where is a covariance matrix and is a standard Wiener process over the state space of . may be a diagonal matrix of variances ( , and ) corresponding to each component of the state vector . Typically , we assume that the above model is coupled to a measurement “device” , which permits indirect observations of the hidden state : ( 6 ) where is an observation noise vector . In the simplest case , the vector of observations is one-dimensional and it may consist of noisy measurements of the membrane potential: ( 7 ) where is the standard deviation of the observation noise and a random number sampled from a Gaussian distribution with zero mean and standard deviation equal to unity . More complicated non-linear , non-Gaussian observation functions may be used when , for example , the measurements are recordings of the intracellular calcium concentration , simultaneous recordings of the membrane potential and the intracellular calcium concentration or simultaneous recordings of the membrane potential from multiple sites ( e . g . soma and dendrites ) of a neuron . Assuming that time is partitioned in a very large number of time steps , such that and the corresponding states are , we can approximate the solution to Eq . 5 using the following difference equation: ( 8 ) where and is a random vector with components sampled from a normal distribution with zero mean and unit variance . The above expression implements a simple rule for computing the membrane potential , activation and inactivation variables at each point of the discretized time based on information at the previous time point and it can be considered as a specific instantiation of the Euler-Maruyama method for the numerical solution of Stochastic Differential Equations [33] . Then , the observation model becomes: ( 9 ) In general , measurements do not take place at every point of the discretized time , but rather at intervals of time steps ( depending on the resolution of the measurement device ) , thus generating a total of measurements . For simplicity in the above description , we have assumed that . However , all the models we consider in the Results section assume . In terms of probability density functions , the non-linear state-space model defined by Eqs . 8 and 9 ( known as the dynamics model ) and the observation model , respectively ) can be written as: ( 10 ) ( 11 ) where the initial state is distributed according to a prior density . The above formulas are known as the state transition and observation densities , respectively [34] . In many inference problems involving state-space models , a primary concern is the sequential estimation of the following two conditional probability densities [29]: ( a ) and ( b ) , where , i . e . the set of observations ( for example , a sequence of measurements of the membrane potential ) up to the time point . Density ( a ) , known as the filter density , models the distribution of state given all observations up to and including the time point , while density ( b ) , known as the smoother density , models the distribution of state given the whole set of observations up to the final time point . In principle , the filter density can be estimated recursively at each time point using Bayes' rule appropriately [29]: ( 12 ) where and are the state transition and observation densities , respectively , and is the filter density at the previous time step . Then , the smoother density can be obtained by using the following general recursive formula: ( 13 ) which evolves backwards in time and makes use of the pre-calculated filter , . Given either of the above posterior densities , we can compute the expectation of any useful function of the hidden model state as: ( 14 ) where is either the filter or the smoother density . Common examples of are itself ( giving the mean ) and the squared difference from the mean ( giving the covariance of ) . In practice , the computations defined by the above formulas can be performed analytically only for linear Gaussian models using the Kalman smoother/filter and for finite state-space hidden Markov models . For non-linear models , the extended Kalman filter is a popular approach , which however can fail when non-Gaussian or multimodal density functions are involved [34] . A more generally applicable , albeit computationally more intensive approach , approximates the filter and smoother densities using Sequential Monte Carlo ( SMC ) methods , also known as particle filters [34] , [35] . Within the SMC framework , the filter density at each time point is approximated by a large number of discrete samples or particles , , and associated non-negative importance weights , : ( 15 ) where is the Dirac delta function centered at the particle , . Given an initial set of particles sampled from a prior distribution and their associated weights , a simple update rule involves the following steps [29]: Step 1: For , sample a new set of particles from the proposal transition density function , . In general , one has enormous freedom in choosing the form of this density and even condition it on future observations , if these are available ( see , for example , [36] ) . However , the simplest ( and a quite common ) choice is to use the transition density as the proposal , i . e . . This is the approach we follow in this paper . Step 2: For each new particle , evaluate the importance weight: ( 16 ) Notice that when , then the computation of the importance weights is significantly simplified , i . e . . Step 3: Normalize the computed importance weights , by dividing each of them with their sum , i . e . ( 17 ) The derived set of weighted samples is considered an approximation of the filter density . In practice , the above algorithm is augmented with a re-sampling step ( preceding Step 1 ) , during which particles are sampled from the set of weighted particles computed at the previous iteration with probabilities proportional to their weights [34] , [35] . All re-sampled particles are given weights equal to . This step results in discarding particles with small weights and multiplying particles with large weights , thus compensating for the gradual degeneration of the particle filter i . e . the situation where all particles but one have weights equal to zero . For performance reasons , the resampling step may be applied only when the effective number of particles drops below a threshold value , e . g . . An estimation of the effective number of particles is given by ( 18 ) The above filter can be extended to a fixed-lag smoother , if instead of resampling just the particles at the current time step , we store and resample all particles up to time steps before the current time step , i . e . [29] . The resampled particles can be considered a realization from a posterior density , which is an approximation of the smoother density , for sufficiently large values of . Within this Monte Carlo framework , the expectation in Eq . 14 can be approximated as: ( 19 ) for a large number of weighted samples . It is possible to apply the above standard filtering and smoothing techniques to parameter estimation problems involving state-space models . The key idea [29] is to define an extended state vector by augmenting the state vector with the model parameters , i . e . . Then , the time evolution of the extended state-space model becomes: ( 20 ) while the observational model remains unaltered: ( 21 ) The marginal posterior density of the parameter vector is given by: ( 22 ) and , subsequently , the expectation of any function of can be computed as in Eq . 14: ( 23 ) Furthermore , given a set of particles and associated weights , which approximate the smoother density as outlined in the previous section , i . e . for , the above expectation can be approximated as: ( 24 ) for large . Under this formulation , parameter estimation , which is traditionally treated as an optimization problem , is reduced to an integration problem , which can be tackled using filtering and smoothing methodologies for state-space models , a well-studied subject in the field of Computational Statistics . It should be emphasized that although in Eq . 20 the parameter vector was assumed constant , i . e . , the same methodology applies in the case of parameters that are naturally evolving in time , such as a time-varying externally injected current . A particularly interesting case arises when an artificial evolution rule is imposed on a parameter vector , which is otherwise constant by definition . Such a rule allows sampling new parameter vectors based on samples at the previous time step , i . e . , and generating a sequence , which explores the parameter space and , ideally converges in a small optimal subset of it , after a sufficiently large number of iterations . It is at this point that the opportunity to use techniques borrowed from the domain of Evolutionary Algorithms arises . Here , we assume that the artificial evolution of the parameter vector is governed by a version of the Covariance Matrix Adaptation algorithm [30] , a well-known Evolution Strategy , although the modeler is free to make other choices ( e . g . Differential Evolution [31] ) . For the particle , we write: ( 25 ) where is a random vector with elements sampled from a normal distribution with zero mean and unit variance . and are a mean vector and covariance matrix respectively , which are computed as follows: ( 26 ) ( 27 ) In the above expressions , a and are small adaptation constants and and are the expectation and covariance of the weighted sample of , respectively . is a scale parameter that evolves according to a log-normal update rule: ( 28 ) where is a small adaptation constant and is a normally distributed random number with zero mean and unit variance . According to Eq . 25 , the parameter vector is sampled at each iteration of the algorithm from a multivariate normal distribution , which is centered at and has a covariance matrix equal to : ( 29 ) Both and are slowly adapting to the sample mean and covariance , with an adaptation rate determined by the constants and . Notice that by switching off the adaptation process ( i . e . by setting ) , evolves according to a multivariate Gaussian distribution , which is centered at the previous parameter vector and has a covariance matrix equal to : ( 30 ) Therefore , given an initial set of weighted particles sampled from some prior density function and an initial covariance matrix , which may be set equal to the identity matrix , the smoothing algorithm presented earlier becomes: Step 1a: Compute the expectation and covariance of the weighted sample of Step 1b: For , compute the scale factor according to Eq . 28 . Notice that this scale factor is now part of the extended state for each particle Step 1c: For , compute the mean vector , as shown in Eq . 26 Step 1d: Compute the covariance matrix , as shown in Eq . 27 Step 1e: For , sample , as shown in Eq . 25 Step 1f: For , sample a new set of state vectors from the proposal density , thus completing sampling the extended vectors . Notice that the proposal density is conditioned on the updated parameter vector . Step 2–3: Execute steps 2 and 3 as described previously Notice that in the algorithm outlined above , the order in which the components of are sampled is important . First , we sample the scaling factor . Then , we sample the parameter vector given the updated . Finally , we sample the state vector from a proposal , which is conditioned on the updated parameter vector . When resampling occurs , the state vectors with large importance weights are selected and multiplied with high probability along with their associated parameter vectors and scaling factors , thus resulting in a gradual self-adaptation process . This self-adaptation mechanism is very common in the Evolution Strategies literature . The algorithm described in the previous section was implemented in MATLAB and C ( source code available as Supplementary Material; unmaintained FORTRAN code is also available upon request from the first author ) and tested on parameter inference problems using simulated or actual electrophysiological data and a number of Hodgkin-Huxley-type models: ( a ) a single-compartment model ( derived from the classic Hodgkin-Huxley model of neural excitability ) containing a leakage , transient sodium and delayed rectifier potassium current , ( b ) a two-compartment model of a cat spinal motoneuron [37] and ( c ) a model of a B4 motoneuron in the Central Nervous System of the pond snail Lymnaea stagnalis [38] , which was developed as part of this study . Each of these models is described in detail in the Results section . Models ( a ) and ( b ) were used for generating noisy voltage traces at a sampling rate of ( one sample every ) . The simulated data was subsequently used as input to the algorithm in order to estimate a large number of parameters; typically , maximal conductances of ionic currents , reversal potentials , the parameters governing the activation and inactivation kinetics of ionic currents , as well as the levels of intrinsic and observation noise . Estimated parameter values were subsequently compared against the true parameter values in the model . The MATLAB environment was used for visualization and analysis of simulation results . For the estimation of the unknown parameters in model ( c ) , actual electrophysiological data were used , as described in the next section . Prior information was incorporated in the smoother by assuming that parameter values were not allowed to exceed well-defined upper or lower limits ( see Tables 1 , 2 and 3 ) . For example , maximal conductances never received negative values , while time constants were always larger than zero . At the beginning of each simulation , the initial population of particles was uniformly sampled from within the acceptable range of parameter values and , during each simulation , parameters were forced to remain within their pre-defined limits . All simulations were performed on an Intel dual-core i5 processor with 4 GB of memory running Ubuntu Linux . The number of particles used in each simulation was typically , where was the dimensionality of the extended state ( equal to the number of free parameters and dynamic states in the model ) . The time step in the Euler-Maruyama method was set equal to . The parameter of the fixed-lag smoother was set equal to ( unless stated otherwise ) , which is equivalent to a time window wide ( since data were sampled every ) . The adaptation constants , and in Eqs . 26 , 27 and 28 were all set equal to , unless stated otherwise . Depending on the size of , the complexity of the model and the length of the ( actual or simulated ) electrophysiological recordings , simulation times ranged from a few minutes up to more than hours . As part of this study , we developed a single-compartment Hodgkin-Huxley-type model of a B4 neuron in the pond-snail Lymnaea stagnalis [38] . B4 neurons are part of the neural circuit that controls the rhythmic movements of the feeding muscles via which the animal captures and ingests its food . The Lymnaea central nervous system was dissected from adult animals ( shell length ) that were bred at the University of Leicester as described previously [39] . All dissections were carried out in -buffered saline containing ( in ) , , , , and , , in distilled water . All chemicals were purchased from Sigma . The buccal ganglia containing the B4 neurons were separated from the rest of the nervous system by cutting the cerebral buccal connectives and the buccal-buccal connective was crushed to eliminate electrical coupling between B4 neurons in the left and right buccal ganglion . Prior to recording , excess saline was removed from the dish and small crystals of protease type XIV were placed directly on top of the buccal ganglia to soften the connective tissue and aid the impalement of individual neurons . The protease crystals were washed of after about with multiple changes of -buffered saline . The B4 neuron was visually identified based on its size and position and impaled with two sharp intracellular electrodes filled with a mixture of potassium acetate and potassium chloride ( resistance ) . During the recording , the preparation was bathed in -buffered saline plus hexamethonium chloride to block cholinergic synaptic inputs and suppress spontaneous fictive feeding activity . The signals from the two intracellular electrodes were amplified using a Multiclamp 900A amplifier ( Molecular Devices ) , digitized at a sampling frequency of using a CED1401plus A/D converter ( Cambridge Electronic Devices ) and recorded on a PC using Spike2 version 6 software ( Cambridge Electronic Devices ) . A custom set of instructions using the Spike2 scripting language was used to generate sequences of current pulses consisting of individual random steps ranging in amplitude from to and a duration from to . The current signal was injected through one of the recording electrodes whilst the second electrode was used to measure the resulting changes in membrane potential .
The applicability of the fixed-lag smoother presented above was demonstrated on a range of Hodgkin-Huxley-type models using simulated or actual electrophysiological data . The first model we examined consisted of a single compartment containing leakage , sodium and potassium currents , as shown below: ( 32 ) where . Notice the absence of noise in the dynamics of , and , which is valid if we assume a very large number of channels ( see Supplementary Material and Supplementary Figures S1 and S2 for the case were noise is present in the dynamics of these variables ) . The steady states and relaxation times of the activation and inactivation gating variables were voltage-dependent , as shown below ( e . g . [5] ) : ( 33 ) and ( 34 ) where and . The parameters , , , and in Eqs . 33 and 34 were chosen such that and fit closely the corresponding steady-states and relaxation times of the classic Hodgkin-Huxley model of neural excitability in the giant squid axon [40] . Observations consisted of noisy measurements of the membrane potential , as shown in Eq . 7 . The full set of parameter values in the above model is given in Table 1 . First , we used the fixed-lag smoother to simultaneously infer the hidden states ( , , , ) and standard deviations of the intrinsic ( ) and observation ( ) noise based on -long simulated recordings of the membrane potential . These recordings were generated by assuming a time-dependent in Eq . 31 , which consisted of a sequence of current steps with amplitude randomly distributed between and and random duration up to a maximum of . Two simulated voltage recordings were generated corresponding to two different levels of observation noise , and , respectively . The second value ( ) was rather extreme and it was chosen in order to illustrate the applicability of the method even at very high levels of observation noise . Simulated data points were sampled every ( ) . The standard deviation of the intrinsic noise was set at . The injected current and the induced voltage trace ( for either value of ) were then used as input to the smoother , during the inference phase . At this stage , all other parameters in the model ( conductances , reversal potentials , and ionic current kinetics ) were assumed known , thus the extended state vector took the form , where was a scale factor as in Eq . 25 . New samples for were taken from a log-normal distribution ( Eq . 28 ) , while new samples for and were drawn from an adaptive bivariate Gaussian distribution at each iteration of the algorithm ( Eq . 25 ) . For each data set , smoothing was repeated for two different values of the smoothing lag , i . e . and . corresponds to filtering , while corresponds to smoothing with a fixed lag equal to . Our results from this set of simulations are summarized in Fig . 1 . We observed that at low levels of observation noise ( Fig . 1A ) , the inferred expectation of the voltage ( solid blue and red lines ) closely matched the underlying ( true ) signal ( solid black line ) . This was true for both values of the fixed lag used for smoothing . However , at high levels of observation noise ( Fig . 1Bi ) , the true voltage was inferred with high fidelity when a large value of the fixed lag ( ) was used ( solid red line ) , but not when ( solid blue line ) . Furthermore , the inferred expectations of the unobserved dynamic variables , and ( solid red lines in Fig . 1Bii ) also matched the true hidden time series ( solid black lines in the same figure ) remarkably well , when . We repeat that during these simulations an artificial update rule was imposed on the two free standard deviations and , as shown in Eq . 25 . The artificial evolution of these parameters is illustrated in Fig . 1Ci , where the inferred expectations of and are presented as functions of time . These expectations converged immediately , fluctuating around the true values of and ( dashed lines in Fig . 1Ci ) . This is also illustrated by the histograms in Fig . 1Cii , which were constructed from the data points in Fig . 1Ci . We observed that the peaks of these histograms were located quite closely to the true values of and ( dashed lines in Fig . 1Cii ) . In summary , the fixed-lag smoother was able to recover the hidden states and standard deviations of the intrinsic and observation noise in the model based on noisy observations of the membrane potential . This was true even at high levels of observation noise , subject to the condition that a sufficiently large smoothing lag was adopted during the simulation . Next , we treated two more parameters in the model as unknown , i . e . the maximal conductances of the transient sodium ( ) and delayed rectifier potassium ( ) currents . The extended state vector , thus , took the form . As in the previous section , new samples for were drawn from a log-normal distribution ( Eq . 28 ) , while , , and were sampled by default from an adaptive multivariate Gaussian distribution at each iteration of the algorithm ( Eq . 25 ) . In order to examine the effect of this adaptive sampling approach on the variance of the inferred parameter distributions , we repeated fixed-lag smoothing on -long simulated recordings of the membrane potential assuming each time that different aspects of the adaptive sampling process were switched off , as illustrated in Fig . 2 . First , we assumed that no adaptation was imposed on or the “unknown” noise parameters and maximal conductances , i . e . the constants , and in Eqs . 26–28 were all set equal to zero . In this case , the multivariate Gaussian distribution from which new samples of , , and were drawn from reduced to Eq . 30 . In addition , we assumed that in the same equation was equal to , for all samples . Under these conditions , the true values of the free parameters were correctly estimated through application of the fixed-lag smoother , as illustrated for the case of and in Figs . 2Ai and 2Aii . Subsequently , we repeated smoothing assuming that the scale factor evolved according to the log-normal update rule given by Eq . 28 with , while and were again set equal to . As illustrated in Figs . 2Bi and 2Bii for parameters and , by imposing this simple adaptation rule on the multivariate Gaussian distribution from which the free parameters in the model were sampled , we managed again to estimate correctly their values , but this time the variance of the inferred parameter distributions ( the width of the histograms in Fig . 2Bii ) was drastically reduced . By further letting the mean and covariance of the proposal Gaussian distribution in Eq . 25 adapt ( by setting in Eqs . 26 and 27 ) , we achieved a further decrease in the spread of the inferred parameter distributions ( Figs . 2C and 2D ) . Parameters and and the hidden states , , and were also inferred with very high fidelity in all cases ( as in Fig . 1 ) , but the variance of the estimated posteriors for and followed the same pattern as the variance of and . It is worth observing that when all three adaptation processes were switched on ( i . e . ) , the algorithm converged to a single point in parameter space within the first of simulation , which coincided with the true parameter values in the model ( see Fig . 2D for the case of and ) . At this point , the covariance matrix became very small ( i . e . all its elements were less than , although the matrix itself remained non-singular ) and the mean was very close to the true parameter vector . We note that and , where stands for the expectation computed over the population of particles . In this case , it is not strictly correct to claim that the chains in Fig . 2Di approximate the posteriors of the unknown parameters and ; since repeating the simulation many times would result in convergence at slightly different points clustered tightly around the true parameter values , it would be more reasonable to claim that these optimal points are random samples from the posterior parameter distribution and they can be treated as estimates of its mode . Depending on the situation , one may wish to estimate the full posteriors of the unknown parameters or just an optimal set of parameter values , which can be used in a subsequent predictive simulation . In Fig . 3A , we examined in more detail how the scale factor affects the variance of the final estimates , assuming that . We repeat that each particle contains as a component of its extended state . Each scaling factor is updated at each iteration of the algorithm following a lognormal rule ( Eq . 28 , Step 1b of the algorithm in the Methods section ) . Sampling new parameter vectors is conditioned on these updated scaling factors ( Eq . 25 , Step 1e of the algorithm ) . When at a later stage weighting ( and resampling ) of the particles occurs , the scaling factors that are associated with high-weight parameters and hidden states are likely to survive into subsequent iterations ( or “generations” ) of the algorithm . During the course of this adaptive process , the scaling factors are allowed to fluctuate only within predefines limits , similarly to the other components of the extended state vector . In Fig . 3Ai , we demonstrate the case where the scaling factors were allowed to take values from the prior interval . We observed that during the course of the simulation ( which utilized -long simulated membrane potential recordings ) , the average value of the scaling factor , , decreased gradually towards and this was accompanied by a dramatic decrease in the variance of the inferred parameters and , which eventually “collapsed” to a point in parameter space located very close to their true values . This situation was the same as the one illustrated in Fig . 2D . Notice that although decreased towards zero , it never actually took this value; it merely became very small ( ) . When we used a prior interval for with non-zero lower bound ( i . e . ]; see Fig . 3Aii ) , the final estimates had a larger variance , providing an approximation of the full posteriors of the “unknown” parameters and . Thus , controlling the lower bound of the prior interval for the scaling factors provides a simple method for controlling the variance of the final estimates . Notice that the variance of the final estimates also depends on the number of particles ( Fig . 3B ) . A smaller number of particles resulted in a larger variance of the estimates ( compare Fig . 3Bi to Fig . 3Bii ) . However , when a large number of particles was already in use , further increasing their number did not significantly affect the variance of the estimates or the rate of convergence ( compare Fig . 3Bii to Fig . 3Aii ) , indicating the presence of a ceiling effect . The adaptive sampling of the scaling factors further depends on parameter in Eq . 28 , which determines the width of the lognormal distribution from which new samples are drawn . The value of this parameter provides a simple way to control the rate of convergence of the algorithm; larger values of resulted in faster convergence , when processing -long simulated recordings ( compare Fig . 4A to Fig . 4B ) . The rate of convergence also depends on the number of particles in use ( compare Fig . 4A to Fig . 4C ) , although it is more sensitive to changes in parameter ; dividing the value of by ( Fig . 4B ) had a larger effect on the rate of convergence than dividing the number of particles by ( Fig . 4C ) . In summary , by assuming an adaptive sampling process for the unknown parameters in the model , we managed to achieve a significant reduction in the spread of the inferred posterior distributions of these parameters . Furthermore , adjusting the prior interval and adaptation rate of the scaling factors provides a straightforward way to control the variance of the estimated posteriors and the rate of convergence of the algorithm . Alternatively , we could have set , i . e . set it to the same constant value for all particles and time steps ( as in Fig . 2A ) . However , by permitting to adapt within a predefined interval , we potentially allow this parameter and , thus , the covariance matrices take large values , which in turn would permit the algorithm to escape local optima in the parameter space . For example , the time profiles of in Figs . 3 and 4 indicate that , early during the simulations , this quantity had relatively large values , which were associated with large variances of the posterior parameter estimates . During this initial period , the algorithm has the potential to “jump” away from local optima and towards more optimal regions of the parameter space . One may see , here , a distant analogy to simulated annealing , where a fictitious “temperature” control variable is gradually decreased , thus allowing the system to escape local minima and gradually settle to more optimal regions of the energy landscape . In a subsequent stage , we treated as unknown two more parameters in the model , i . e . the reversal potentials for the sodium and potassium currents , and , respectively . Thus , the extended state vector became . This time , we wanted to examine how increasing levels of observation noise ( i . e . the value of parameter ) affect the inference of unknown quantities in the model based on the fixed-lag smoother . For this reason , we repeated smoothing on four -long simulated data sets ( i . e . recordings of membrane potential and the associated ) corresponding to increasing values of the standard deviation of the observation noise , i . e . , , and . The results from this set of simulations are summarized in Fig . 5 . For , the expectations of the four parameters , , and ( red solid lines in Figs . 5Ai–iv ) eventually converged to their true values ( dashed lines in the aforementioned figures ) . For , the expectations of these parameters ( light red solid lines in Figs . 5Ai–iv ) also converged , although the expectations for ( Fig . 5Ai ) and , to a lesser degree , ( Fig . 5Aii ) deviated noticeably from their true values . As expected , at higher levels of noise , the variance of the final estimates was larger , although the rate of convergence did not seem to be affected , due to the large number of particles we used ( ; see ceiling effect in Fig . 3Bii ) . The inferred parameters and ( not illustrated for clarity ) followed a similar convergence pattern . In Fig . 5B , we show , for each tested value of , the box plots of the above four parameters , which were computed from the data points ( as in Fig . 5A ) corresponding to time . For each parameter and each value of , the data were first normalized as follows: ( 35 ) where . The box plots in Fig . 5B were constructed from the normalized data points . The above normalization was necessary since it made possible the comparison between different data sets , each characterized by its own mean , variance and unit of measurement . In the box plots in Fig . 5B , zero ( i . e . the dashed lines ) corresponds to the true parameter values , while discrepancies from the true parameter values along the y-axis are given in relation to the average . We may observe that for very low levels of observation noise ( ) , the posteriors of the four examined parameters were clustered tightly around their true values , but for larger levels of noise ( , and ) , we observed larger discrepancies from the true parameter values and broader inferred posteriors . The parameters following more noticeably this trend were the conductances and , while and , particularly , were less affected . This indicates that smoothing is more sensitive to changes in some model parameters than others and this is why these parameters were tightly controlled . In summary , increasing the levels of measurement noise ( i . e . the value of parameter ) decreased the accuracy and precision of the algorithm , but it did not significantly affect the rate of convergence due to the large number of particles used during the simulations . At the next stage , we treated all parameters in the model ( a total of 23 parameters; see Table 1 ) as unknown . Therefore , the extended state vector took the following ( -dimensional ) form:where and . These parameters included the standard deviations of intrinsic and observation noise ( and , respectively ) , the maximal conductances and reversal potentials of all currents in the model and the parameters controlling the steady-states and relaxation times of activation and inactivation for the sodium and potassium currents ( , , , and ) . The results from this simulation are illustrated in Fig . 6 . We observed that the true signal ( membrane potential ) was inferred with very high fidelity ( Fig . 6Ai ) . The sodium activation was also recovered with very high accuracy , while estimation of the hidden states and ( sodium inactivation and potassium activation , respectively ) was also satisfactory ( despite significant deviations , the general form of the true hidden states was recovered without any observable impact on the dynamics of the membrane potential ) , as shown in Fig . 6Aii . Among the estimated parameters , we illustrate ( in Figs . 6B and 6C ) the estimated posteriors for the reversal potential of sodium ( Fig . 6B ) and for parameters ( Figs . 6Ci , ii ) and ( Figs . 6Ciii , iv ) , which control the activation of sodium and potassium currents , respectively . We focus on these parameters , because they represent three different characteristic cases . The posteriors of parameters and are unimodal ( see Figs . 6Bii and 6Cii ) and they were estimated with relatively high accuracy . Particularly , the posterior for was estimated with very high precision and accuracy , despite its broad prior interval ( the y-axis in Fig . 6Ci and the x-axis in Fig . 6Cii ) . On the other hand , the estimated posterior of covered a large part of its prior interval ( the y-axis in Fig . 6Ciii and the x-axis in Fig . 6Civ ) , its main mode was located at a slightly larger value than the true parameter value , while at least two minor modes seem to be present near the upper bound of the prior interval ( the arrow in Fig . 6Civ ) . These results reiterate our previous conclusion that smoothing may be particularly sensitive to some parameters , but not to others . The posteriors of parameters in the former category are very precise and narrow ( as in the case of and , especially , ) , while the parameters in the latter category are characterized by broader posteriors . Also , we can observe that the fixed-lag smoother has the capability to provide a global approximation of the unknown posteriors , including their variance and the location of major and minor modes ( i . e . global and local optima ) . An overview of all inferred posteriors is given by the box plot in Fig . 6D , which was constructed after all data ( as in Figs . 6Bi , 6Ci and 6Ciii ) were normalized according to Eq . 35 . Again , it may be observed that while some of the estimated parameter posteriors are quite precise and accurate , such as ( parameter ) , ( parameter ) and ( parameter ) , others are less precise and accurate , such as the maximal conductances ( parameters to ) , ( parameter ) and ( parameter ) . The simulation results presented above were obtained by assuming a prior interval for the scaling factors equal to . When we repeated the simulation using the prior interval , the true underlying membrane potential was again inferred with very high fidelity ( Fig . 7Ai ) , while the hidden states , and were also estimated with sufficient accuracy ( Fig . 7Aii ) . In this case , however , the estimates of the “unknown” parameters converged to single points in parameter space ( as illustrated , for example , for parameters , and in Figs . 7Bi–ii ) , which fall within the support of the posteriors illustrated in Figs . 6B and 6C . The activation and inactivation steady states ( Fig . 7Ci , red solid lines ) and relaxation times ( Fig . 7Cii , red solid lines ) as functions of voltage , which were computed from these estimates , were also similar to their corresponding true functions , with the curves for and manifesting the largest deviation from truth ( black solid lines in Figs . 7Ci , ii ) . An overview of the estimated parameter values ( after normalizing using Eq . 35 ) is given in Fig . 7Di . As stated previously , some estimates were close to their true counterparts , while others were not . For example , the activation of the sodium current ( Fig . 7Aii ) and its steady state ( Fig . 7Ci ) , which are important for the correct onset of the action potentials , were inferred with relatively high accuracy . On the other hand , larger errors were observed , for example , in the inference of sodium inactivation ( ; Fig . 7Aii ) or in the estimation of ( parameter ; Fig . 7Di ) , the maximal conductance for the sodium current . Given the fact that the data on which inference was based ( a single noisy recording of the membrane potential ) was of much lower dimensionality than the extended state we aimed to infer , the observed discrepancies between inferred and true model quantities were unlikely to vanish unless we imposed more strict constraints on the model . When we repeated the previous simulation using more narrow prior intervals for some of the parameters controlling the kinetics of the sodium and potassium currents in the model ( see red dashed boxes in Fig . 7Dii and bold intervals in Table 1 ) , the estimated parameters settled closer to their true values ( Fig . 7Dii ) . This was true even for parameters on which more narrow intervals were not directly applied , such as the maximal conductances ( i . e . parameters to in Fig . 7Dii ) , and even when data with higher levels of observation noise were used ( Fig . 7Dii , data points indicated with crosses; see also Fig . S3 ) . It is important to mention that using more narrow prior constraints only affected the accuracy of the final estimates , not the quality of fitting the experimental data , which in all cases was of very high fidelity . Alternatively , we could have constrained the model by increasing the dimensionality of the observed signal , e . g . by using simultaneously more that one unique voltage traces ( each generated under different conditions of injected current ) during smoothing . We examine the use of multiple data sets simultaneously as input to the fixed-lag smoother later in the Results section . In summary , the smoothing algorithm can be used to resolve high-dimensional inference problems . In combination with sufficient prior information ( in the form of bounded regions within which parameters are allowed to fluctuate; see Table 1 ) , the fixed-lag smoother can provide estimates of the intrinsic and observation noise , maximal conductances , reversal potentials and kinetics of ionic currents in a single-compartment Hodgkin-Huxley-type neuron model , based on low-dimensional noisy experimental data . Next , we tested whether the fixed-lag smoother could be successfully applied on inference problems involving more complex models than the one we used in the previous sections . For this reason , we focused on a two-compartment model of a vertebrate motoneuron containing sodium , potassium and calcium currents and intracellular calcium dynamics , which were differentially distributed among a soma and a dendritic compartment [37] . The model ( modified appropriately to include intrinsic noise terms ) is summarized below: ( 36 ) ( 37 ) where and is the membrane potential at the soma and dendritic compartments , respectively , and . The leakage conductance and reversal potential were and , respectively . The coupling conductance was and the ratio of the soma area to the total surface area of the cell was . The various ionic currents in the above model were as follows: ( a ) a transient sodium current , , ( b ) a delayed rectifier potassium current , , ( c ) a calcium-activated potassium current , , where and ( the half-saturation constant ) , ( d ) an N-type calcium current , , where and ( e ) an L-type calcium current , . The various activation and inactivation dynamic variables in the above model were modeled using first-order relaxation kinetics ( as in Eq . 32 ) , where the various steady states were assumed to be sigmoid functions of voltage ( Eq . 33 ) . Notice , that the activation of was assumed instantaneous and therefore , it was given at any time by the voltage-dependent steady state . The relaxation times for sodium inactivation and potassium activation were also functions of voltage as in Eq . 34: ( 38 ) ( 39 ) where the parameters , and ( with and ) were chosen by fitting the above expressions to the original model in [37] . The relaxation times for the remaining activation and inactivation variables were constant . All parameters values in the model are given in Table 2 . The intracellular calcium concentration at either the soma or the dendritic compartment was also modeled by a first-order differential equation , as follows: ( 40 ) where , and . The total calcium current is at the soma ( ) and at the dendritic compartment ( ) . The observation model assumed simultaneous noisy recordings of the membrane potential from both the soma and dendritic compartments , as follows: ( 41 ) where with . Notice that is the same for both compartments . In the above model , the externally injected currents and were sequences of random current steps with duration up to ( instead of as in the single-compartment model , due to the presence of slower currents in the two-compartment model ) and magnitude between and . Current was injected in both the dendritic compartment and the soma ( instead of just in the soma ) , because preliminary simulations indicated that this experimental setting facilitated parameter estimation , presumably due to the generation of a more variable ( and , thus , information-rich ) data set . The injected currents and the induced noisy voltage traces and comprised the simulated data on which parameter estimation was based . First , we aimed to infer the noise parameters and maximal conductances of all voltage- and calcium-gated currents in the model , assuming that the kinetics of these currents were known . This implied an extended-state vector with components as shown belowwhere . The results from this simulation are illustrated in Figs . 8 and 9 . The fixed-lag smoother managed to recover the hidden dynamic states ( including the time-evolution of the intracellular calcium; Fig . 8 ) , the standard deviations of the intrinsic and observation noise ( Figs . 9Ai , ii ) and the true values of all the gated maximal conductances ( Figs . 9Bi–iv ) in the model using approximately of simulated data and particles . Notice that , in Figs . 8Ci–iv , the inferred hidden gating states ( dashed red lines ) coincide extremely well with the true ones ( solid black lines ) , which is not surprising , since the voltage-dependent kinetics of these states were assumed known and the true membrane potential at the soma and dendritic compartment was recovered with very high fidelity ( Figs . 8Ai , ii ) . Also , notice that , in Figs . 9Aii , 9Biii and 9Biv , the estimation of the standard deviation of the intrinsic noise , , and the maximal conductances of calcium and calcium-dependent currents in the dendritic compartment ( , and ) was improved after injecting current in both the soma and the dendritic compartment ( compare the grey solid lines , which correspond to injection in the soma only , to the color ones in the aforementioned figures ) . In a second stage , we assumed that the kinetics of all voltage-gated ionic currents were also unknown , implying an extended state vector with components , as follows:where , and . Our results from this simulation are summarized in Figs . 10 and 11 . Again , the membrane potential at the soma and the dendrite were inferred with very high fidelity ( Fig . 10Ai , ii ) . However , the estimated hidden dynamics of most ionic currents and intracellular calcium concentrations in the model deviated significantly from their true counterparts ( Fig . 10B , C ) . The expectations of all estimated parameters are illustrated in Fig . 11Ai . As in the case of the single-compartment model , by imposing tighter prior constraints on some of the parameters controlling the kinetics of ionic currents in the model ( see red dashed box in Fig . 11Aii and Table 2 ) , we managed to reduce the discrepancies of the estimates from their true values ( Fig . 11Aii and Supplementary Fig . S4 ) . This was true even for parameters on which stricter priors were not directly applied . The inference was completed after processing almost of data , as shown in Fig . 11B for the maximal conductances of sodium and potassium currents at the soma . Interestingly , the algorithm seems to temporarily settle at local optima ( see arrows in Fig . 11B ) before “jumping” away and , eventually , converge at the final estimates . The inferred voltage-dependent steady-states of the sodium , potassium and calcium currents ( Figs . 11Ci , ii ) and the relaxation times for the sodium inactivation and potassium activation ( Fig . 11Ciii ) were also very similar to their true corresponding functions . The algorithm remained operational when more noisy data were used , as illustrated in Fig . 11Aii and in Supplementary Fig . S5 . An interesting fact regarding the simulation results presented in Figs . 10 and 11Ai was that , in order to obtain high-fidelity estimates of the true membrane potential at the soma and dendritic compartment ( as shown in Figs . 10Ai , ii ) we had to use more than particles , the number calculated by the rule ( see Methods ) . In particular , we used particles , although we cannot exclude that a smaller number may have sufficed . After applying more narrow prior constraints ( Figs . 11Aii , B , C , S4 and S5 ) , using the number of particles calculated by the above simple heuristic ( in this case ) was again sufficient for accurately inferring the true membrane potential ( see Fig . S4Ai , ii and S5Ai , ii ) . This implies that as the complexity ( and dimensionality ) of the estimation problem increases , a non-linearly growing number of particles may be required in order to obtain acceptable results , but this situation may be compensated for by providing highly informative priors . It should be mentioned that the two-compartment model allows for the physical separation of currents and as such it is a slightly better approximation of a real neuron with differential expression of individual currents in different cellular compartments . However , in no way does it capture the full morphological complexity of a real neuron . As such , current injection into the dendritic compartment can not be replicated accurately in a real neuron , as current injection in the model will have a uniform effect on all currents in that compartment , whilst current injection into the dendrite of a neuron would have far more complex effects on dendritic currents , which potentially would be dependent on the distance from the injection site . Thus , whilst it would be possible , albeit challenging , to carry out dual recordings from the soma and dendrites in a real neuron this would not be the same as the dual current injection in the model . In this case , application of the fixed-lag smoother on a more spatially detailed model would be necessary ( and feasible ) . In principle , the method can also assimilate other types of spatial data , such as calcium imaging data , in case recordings from multiple neuron locations are not available ( although we do not examine this case in detail in this paper ) . Given the large number of unknown parameters and hidden states in combination with the low dimensionality of the data ( notice that the intracellular calcium concentration was assumed unobserved ) , it was truly remarkable that the algorithm managed to recover much of the extended state vector with relatively satisfactory accuracy . However , it should be noted that in our simulations we assumed knowledge of important information , such as the passive conductances and and the reversal potentials of sodium , potassium and calcium currents . This and the fact that the availability of prior information in the form of more narrow parameter boundaries improved significantly the accuracy of the final estimates emphasizes our previous conclusion that prior information is important for the successful inference of unknown model parameters and hidden model states using the fixed-lag smoother . Given such information , inference in complex compartmental models based on simultaneous recordings from several neuron locations and , possibly , measurements of intracellular calcium , can be naturally achieved via appropriate formulation of the extended state vector and application of the fixed-lag smoother . In a final set of simulations , we applied the smoother on actual electrophysiological data in order to estimate the unknown parameters in a single-compartment model of the B4 motoneuron from the nervous system of the pond snail , Lymnaea stagnalis [38] . This neuron is part of a population of motoneurons , which receive rhythmic electrical input from upstream Central Pattern Generator interneurons and in turn innervate and control the movements of the feeding muscles via which the animal captures and ingests its food . Previous studies in these neurons have demonstrated the presence of a transient inward sodium current , a delayed outward potassium current and a transient outward potassium current [41] . A hyperpolarization-activated current was conditional on the presence of serotonin in the solution [38] and , therefore , this current was not included in this instance of the B4 model . Thus , the current conservation equation for a single-compartment model of the B4 motoneuron ( appropriately modified to include an intrinsic noise term ) took the following form: ( 42 ) where the leakage conductance , leakage reversal potential and membrane capacitance in the above model were estimated a priori based on neuron responses to negative current pulses ( , and , respectively ) . The voltage-activated currents that appear in the above expression were modeled as follows: ( a ) , ( b ) and ( c ) , where and as in [41] . The dynamic activation and inactivation variables of these currents ( , , and ) obeyed first-order relaxation kinetics ( as in Eq . 32 ) with voltage-dependent steady-states ( Eq . 33 ) and relaxation times ( Eq . 34 with and ) , similarly to previously published neuron models in the central nervous system of Lymnaea [42] . The observation model was as in Eq . 7 . The raw data we used for inferring the parameters in the above model took the form of four independent -long recordings of the membrane potential from the same B4 motoneuron . Each recording was taken while injecting an external current in the neuron consisting of a sequence of random steps ranging in amplitude between and and with duration between and . A particular characteristic of the data generated under these conditions was the presence of brief bursts of spikes , which were interrupted by relatively long intervals of non-activity ( corresponding to sub-threshold excitatory and inhibitory current injections , respectively; see Figs . 12Ai–iv ) . These long intervals of inactivity were not informative and they negatively affected the performance of the smoother by permitting the random drift of particles towards non-optimal regions of the parameter space ( see Supplementary Fig . S6 ) . However , when the four recordings are considered together , the intervals of inactivity at any single voltage trace overlap with intervals of activity at the remaining three voltage traces , resulting in a four-dimensional data set , where the overall intervals of inactivity were minimized . This four-dimensional data set was used as input to the smoother during the inference phase . Thus , the -dimensional extended state vector became:where , and . Notice the presence of four groups of hidden dynamic states , { , , , , , } , where each group corresponds to a different voltage trace ( and associated externally injected current , ) . The evolution of all four groups of dynamic variables was governed by a common ( shared ) set of parameters . In total , we had to estimate unknown parameters . The boundaries within which the values of these parameters were allowed to fluctuate are given in Table 3 ( indicated in bold ) and they were chosen from within the support of the posteriors in Supplementary Fig . S7 ( after a few trial-and-error simulations ) , which were obtained by using the broader prior intervals given in Table 3 . Notice that the marginal distributions illustrated in Fig . S7 have large variance and multiple modes and , although they provide a global view of the structure of the parameter space , they cannot be used to identify a single combination of optimal parameters values , since they do not include any information regarding correlations between parameters . Using the major modes of the inferred posteriors did not lead to an accurate ( or even spiking ) predictive model . Thus , the estimation was based on using more narrow prior intervals , which helped us estimate unimodal posteriors with small variance ( see Fig . 12C ) and , thus , identify a single combination of optimal parameters that could be used in predictive simulations . We cannot prove that other optimal combinations of parameters do not exist , but we were not able to find any ( i . e . by choosing different narrow prior intervals ) after a reasonable amount of time . Also , notice that the standard deviations of the intrinsic and observation noise were not subject to estimation , but instead they were given ( through trial and error ) the minimal fixed values and , respectively . If left free during smoothing , the values of these parameters fluctuated uncontrollably , masking the contribution of the remaining parameters in the model and , thus , achieving an almost perfect ( but meaningless ) smoothing of the experimental data . This is an indication that the B4 model we used may be missing one or more relevant components , such as additional currents and compartments ( see below for further analysis of this point ) . We did not observe this effect in the cases examined in the previous sections , where simulated data was used , because the models responsible for the generation of this data were , by definition , precisely known . Our results from this set of simulations are illustrated in Fig . 12 . Simultaneous smoothing of all four data sets was again accomplished with high fidelity , as illustrated in Figs . 12Ai–iv . The artificial evolution of the expectations of the conductances for the transient sodium , persistent potassium and transient potassium currents , as well as of some of the kinetic parameters that were estimated in the model is illustrated in Figs . 12Bi–iii . The distributions of all inferred parameters ( normalized after replacing in Eq . 35 with , for each tested parameter ) are also illustrated in Fig . 12C . The inferred expectations of all parameters are given in Table 3 . In order to examine the predictive value of the model given the estimated parameter expectations in Table 3 , we compared its activity to that of the biological B4 neuron , when both were injected with a -long random current consisting of a sequence of current pulses with amplitude ranging from to and duration from to . Our results from this simulation are illustrated in Fig . 13 . We observed that the overall pattern of activity of the model was similar to that of the biological neuron ( Fig . 13A ) . Whilst the model overall generated more action potentials , some individual spikes were absent in the simulated data . A more detailed examination of our data revealed specific differences between the biological and model neurons , which explain the differences in the overall activity between the two ( Fig . 13B , C ) . The spike shape of the model neuron was quite similar to that of its biological counterpart ( Fig . 13Bi ) , including spike threshold , peak , trough and height ( i . e . trough-to-peak amplitude; Fig . 13Biii ) , but the simulated spike had a slightly longer duration than the biological one ( half-width: vs ; Fig . 13Bii ) . In a second set of experiments , both the biological and model neurons were injected with -long current pulses ranging from to and their current-voltage ( IV ) and current-frequency ( IF ) relations were constructed ( Fig . 13C ) . The IV plot showed some non-linear behavior in response to negative current pulses in the experimental data ( probably due to the presence of a residual current ) , which was not present in the simulations ( Fig . 13Ci ) . As a result , the slope of the part of the IV curve corresponding to was more shallow in the simulations than in the experimental data . Moreover , the rheobase was lower in the experimental data than in the model , but the slope of the IF curve was steeper in the simulated data , which resulted in higher firing rates for the model at injected currents larger than approximately ( Fig . 13Cii ) . This feature can account for the overall level of spiking in the model neuron when compared to the biological one ( Fig . 13A ) . Overall , this analysis illustrates that the assumed B4 model did not capture all the aspects of the real neuron . However , this does not mean that our estimation method is flawed . It just shows that the model is actually missing some relevant components , such as additional ionic currents or compartments , which would be necessary for approximating more accurately the spatial structure and biophysical properties of the biological neuron . In the first part of the manuscript we have demonstrated that if the underlying model is complete , then our method produced accurate estimates of the true parameter values , given sufficient informative priors . Thus , it is safe to assume that the observed differences between the biological and model neurons can be minimized , if the fixed-lag smoother is applied on a more complex model of the B4 motoneuron . In summary , we used the fixed-lag smoother to estimate the unknown parameters in a single-compartment model of an invertebrate motoneuron based on actual electrophysiological data . The model , although a simplification of the actual biological system , was still quite complex containing a number of non-linearly interacting components and a total of unknown parameters . By using the methodologies outlined in the previous sections , we managed to estimate the values of these parameters , such that the resulting model mimicked with satisfactory accuracy the overall activity of its biological counterpart . Furthermore , we demonstrated the flexibility of the fixed-lag smoother by showing how it can be used to process simultaneously multiple data sets , given an appropriate formulation of the extended state vector .
Parameter estimation in conductance-based neuron models traditionally involves a global optimization algorithm ( for example , an evolutionary algorithm ) , usually in combination with a local search method ( such as gradient descent ) , in order to find combinations of model parameters that minimize a pre-defined cost function . In this paper , we have addressed the problem of parameter estimation in Hodgkin-Huxley-type models of single neurons from a different perspective . By adopting a hidden-dynamical-systems formalism and expressing parameter estimation as an inference problem in these systems , we made possible the application of a range of well-established inference methods from the field of Computational Statistics . Although it is usually assumed that the kinetics of ionic currents in a conductance-based model are known a priori , here we assumed that this was not the case and , typically , we estimated kinetic parameters , along with the maximal conductances and reversal potentials of ionic currents in the models we examined . The particular method we used was Kitagawa's self-organizing state-space model , which was implemented as a fixed-lag smoother . The smoother was combined with an adaptive algorithm for sampling new sets of parameters akin to the Covariance Matrix Adaptation Evolution Strategy . Alternatively , we could have approximated the smoother distribution ( Eq . 13 ) with a two-pass algorithm , consisting of a forward filter followed by a backward smoothing phase , which would make use of the precomputed filter [34] . This would require storing the filter for the whole duration of the smoothed data , which in turn would have very high memory requirements when large numbers of particles or high-dimensional problems are considered . In contrast , the fixed-lag smoother has the advantage that only the particles up to time steps in the past need to be stored , which is less demanding in memory size and computationally more efficient . Moreover , the fixed-lag smoother , being a single-pass algorithm , was more natural to use in the context of on-line parameter estimation . The applicability of the algorithm was demonstrated on a number of conductance-based models using noisy simulated or actual electrophysiological data . In a recent study , it was found that increasing observation noise led to an increase in the variance of parameter estimates and a decrease in the rate of convergence of the algorithm [28] . Similarly , we observed that at high levels of observation noise , although the algorithm remained functional , its accuracy and precision were reduced ( Fig . 5 ) . It is emphasized that , at a particular level of observation noise , the outcome of the algorithm is an approximation of the posterior distributions of hidden states and unknown parameters in the model , given the available experimental data and prior information . In general , these approximate posteriors provide an overview of the structure of the parameter space and they potentially have multiple modes ( or local optima ) . By taking advantage of the adaptive nature of the fixed-lag smoother ( and , in particular , by controlling the scaling factor that determines the width of the proposal distribution in Eq . 25 ) , we managed to reduce the variance of these posteriors and , in the limit case , we could force the algorithm to converge to a single optimal point ( belonging to the support of the parameter posteriors ) , which could subsequently be used in predictive simulations ( e . g . see Figs . 7D and 11A ) . Unlike the study in [28] , we did not observe any significant reduction in the rate of convergence of the algorithm at high levels of observation noise , which was attributed to a ceiling effect due to the large number of particles we used in our simulations ( typically , , where was the dimensionality of the estimation problem; see Figs . 3B and 4C ) . Thus , we cannot exclude observing such a reduction in the rate of convergence , if a smaller number of particles is used and/or problems of higher dimension are examined . Furthermore , the proposed method requires only a single forward pass of the experimental data , instead of multiple passes , as in the case of off-line estimation methods , including the Expectation Maximization ( EM ) algorithm . On the other hand , this means that , in general , the proposed algorithm requires processing longer data time series in order to converge . In addition , unlike off-line estimation methods , it does not take into account the complete data trace at each iteration , but at most past data points ( but , also , see [36] for a partial “remedy” of this situation ) . In principle , it would be possible to combine previous work on parameter estimation ( e . g . [43] , [44] ) within an EM inference framework in order to estimate various types of parameters ( including maximal conductances and channel kinetics ) in conductance-based neuron models . This could be an interesting topic for further research . Our main conclusion was that , using this algorithm and a set of low-dimensional experimental data ( typically , one or more traces of membrane potential activity ) , it was possible to fit complex compartmental models to this data with high fidelity and , simultaneously , estimate the hidden dynamic states and optimal values of a large number of parameters in these models . Based on simulation experiments using simulated data , we found that the estimated optimal parameter values and hidden states were close to their true counterparts , as long as sufficient prior information was made available to the algorithm . This information took the form of knowledge of the values of particular parameters ( for example , the passive properties of the membrane ) or of relatively narrow ranges of permissible parameter values . Such prior information could have included the kinetics of the ion currents that flow through the membrane or the spatial distribution of various parameter values along different neuron compartments ( e . g . the ratio of maximal conductance A between compartment 1 and compartment 2 ) . In real-life situations , such information may become available through current- or voltage-clamp experiments . For example , the passive properties in the B4 model ( membrane capacitance , leakage maximal conductance and reversal potential ) were inferred from current-clamp data and , thus , they were fixed during the subsequent smoothing phase . It has been demonstrated that this requirement for prior information may be relaxed , if the data set used as input was sufficiently variable to tease apart the relative contribution of different parameters in a model [15] . A well-established result in conductance-based modeling is that the same pattern of electrical activity may be produced by different parameter configurations of the same model [6]–[9] . This implies that it is impossible to identify , during the course of an optimization procedure , a unique set of parameters using just this single pattern of activity as input to the method . For example , as we observed in the case of the B4 model , the posteriors of the estimated parameters may be characterized by multiple modes ( i . e . local optima ) or quite large variances , which makes identification of a unique set of optimal parameter values for use in predictive simulations rather difficult ( Supplementary Fig . S7 ) . A more variable data set would be necessary in order to constrain the model under study , thus forcing the optimization process to converge towards a unique solution . It should be noted that this conclusion was reached by treating as unknown only the maximal conductances in a conductance-based model [15] . Although it is reasonable to assume that this holds true when the kinetics of ion channels are also treated as unknown , it still needs to be demonstrated whether the generation of a data set sufficiently variable to constrain both the maximal conductances and kinetics of ion channels in a complex conductance-based model is practical or even feasible . A more pragmatic approach would be to rely on a mixture of prior information and one or more sufficiently variable electrophysiological recordings as input to the optimization algorithm . It was shown in this study that both the injection of prior information ( in the form mentioned above ) and the simultaneous assimilation of multiple data sets is straightforward using the proposed algorithm . It is important to notice that , unlike more traditional approaches , explicitly defining a cost or fitness function was not required by the fixed-lag smoother . Given the fact that the efficiency of any optimizer can be seriously impeded by a poorly designed cost function , bypassing the need to define such a function may be viewed as an advantage of the proposed method . As in previous studies [43] , [44] , here lies the implicit assumption that by fitting ( or smoothing ) with high fidelity the raw experimental data ( for example , one or more recordings of the membrane potential ) , the estimated model would capture a whole range of features embedded in this data , such as the current-frequency response of the neuron . Although this is a reasonable assumption , we found that it did not hold completely true , when our knowledge of the form of the underlying model was not exact , as in the case of the B4 neuron . In this case , although we could achieve a very good smoothing of the experimental data , subsequent predictive simulations using the inferred model parameters revealed discrepancies between simulation output and experimental data . It is likely that these discrepancies will be minimized , if important missing components are added to the model , such as additional ionic currents or , importantly , an approximation of the spatial structure of the biological neuron . An important outcome of this study was to demonstrate the intimate relation between the self-organizing state-space model and evolutionary algorithms . When used for parameter estimation , the self-organizing state-space model undergoes at each iteration a process of new particle ( individual ) generation ( mutation/recombination ) and resampling ( selection and multiplication ) , which parallels similar processes in evolutionary algorithms . At the root of this parallelism is the fact that we need to impose an artificial evolution on model parameters as part of the formulation of the self-organizing state-space model ( see Methods ) , thus providing a unique opportunity to merge the two classes of algorithms . Here , we decided to combine the self-organizing state-space model with an adaptive algorithm similar to the Covariance Matrix Adaptation Evolution Strategy [30] and by following this adaptive strategy , we managed to achieve a dramatic reduction in the variance of parameter estimates . However , this choice is by no means exclusive and other evolutionary algorithms may be chosen instead , e . g . the Differential Evolution algorithm [31] . This is a topic open to further exploration . Notice that , similarly to Evolutionary Algorithms , the proposed method has , in principle , the ability to estimate the possibly multi-modal posterior distribution of the unknown parameters in the examined model , i . e . it is a global estimation method ( for example , see Fig . 6C , 11B and S7 ) . At each iteration , the algorithm retains a population of particles , which are characterized by a degree of variability and , thus , give the algorithm the opportunity to randomly explore a wide range of the parameter space , spending on average more time in the vicinity of optimal regions . By imposing narrow prior constraints on some of the unknown parameters , we are effectively reducing the dimensionality of the problem and we force the algorithm to converge towards a particular optimum , which can be later used in predictive simulations . A point of potential improvement concerns our choice of the proposal density , . Here , we made the common and straightforward choice to use the transition density as our proposal . However , the modeler is free to make other choices . For example , a recent study demonstrated that the efficiency of particle filters can be significantly increased by conditioning the proposal density on future observations [36] . An important practical aspect of the proposed algorithm was its high computational cost . This cost increased as a function of the number of particles used during smoothing , the length of the fixed smoothing lag , the complexity of the model and the number of unknown parameters in the model . Our simulations on an Intel dual-core i5 processor with four gigabytes of memory took from a few minutes to more than 12 hours to complete . An emerging trend in Scientific Computing is the use of modern massively parallel Graphics Processing Units ( GPUs ) in order to accelerate general purpose computations , as those presented in this paper . The utility of this approach in achieving significant accelerations of Monte Carlo simulations has been recently demonstrated [45] and it has even been applied recently on parameter estimation problems in conductance-based models of single neurons [46] . Preliminary results using a GPU-accelerated version of the fixed-lag smoother ( data not shown ) have indeed demonstrated reduced simulation times , but the accelerations we observed were not as dramatic as those reported in the literature [45] , [46] . This can always be attributed to the fact that our implementation of the algorithm was not optimized . On the other hand , we observed significant accelerations in our simulations involving the serial implementation of the fixed-lag smoother , just by switching from an open-source compiler ( GNU ) to a commercial one ( Intel ) , which presumably emitted better optimized machine code for the underlying hardware . Nevertheless , the use of GPUs for general purpose computing is becoming common and it is likely to become quite popular with the advent of cheaper hardware and , importantly , more flexible and programmer-friendly Application Programming Interfaces ( APIs ) . Overall , our results point towards a generic four-stage heuristic for parameter estimation in conductance-based models of single neurons: ( a ) First , the general structure of the model is decided , such as the number of ionic currents and compartments it should include . ( b ) Second , prior information is exploited in order to fix as many parameters as possible in the model and tightly constrain the remaining ones . For example , the capacitance , reversal potentials and leakage conductance in the model may be fixed to values estimated from current-clamp data . By further exploiting current– and voltage-clamp data , narrow constraints may be imposed on the remaining free ( e . g . kinetic ) parameters in the model . ( c ) At a third stage , more precise parameter value distributions are estimated by applying the fixed-lag smoother on current-clamp data , such as one or more recordings of the electrical activity of the membrane induced by random current injections . ( d ) Finally , the predictive value of the model is assessed through comparison to independent data sets and the model is modified , if necessary . It is important to notice that the techniques outlined in this paper are applicable on a wide range of research domains and that they provide a disciplined way to merge complex stochastic dynamic models , noisy data and prior information under a common inference framework . In conclusion , the class of statistical estimation methods , which the algorithm presented in this paper belongs to , in combination with Monte Carlo approximation techniques are particularly suitable to address high-dimensional inference problems in a disciplined manner . This makes them potentially useful tools at the disposal of biophysical modelers of neurons and neural networks and it is predicted that these methodologies will become more popular in the future among this research community . | Parameter estimation is a problem of central importance and , perhaps , the most laborious task in biophysical modeling of neurons and neural networks . An emerging trend is to treat parameter estimation in this context as yet another statistical inference problem , which can be tackled using well-established methods from Computational Statistics . Inspired by these recent advances , we adopted a self-organizing state-space-model approach augmented with an adaptive sampling algorithm akin to the Covariance Matrix Adaptation Evolution Strategy in order to estimate a large number of parameters in a number of Hodgkin-Huxley-type models of single neurons . Parameter estimation was based on noisy electrophysiological data and involved the maximal conductances , reversal potentials , levels of noise and , unlike most mainstream work , the kinetics of ionic currents in the examined models . Our main conclusion was that parameters in complex , conductance-based neuron models can be inferred using the aforementioned methodology , if sufficiently informative priors regarding the unknown model parameters are available . Importantly , the use of an adaptive algorithm for sampling new parameter vectors significantly reduced the variance of parameter estimates . Flexibility and scalability are additional advantages of the proposed method , which is particularly suited to resolve high-dimensional inference problems . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
]
| [
"mathematics",
"statistics",
"biology",
"computational",
"biology",
"neuroscience",
"probability",
"theory"
]
| 2012 | A Self-Organizing State-Space-Model Approach for Parameter Estimation in Hodgkin-Huxley-Type Models of Single Neurons |
Eukaryotic mRNAs undergo a cycle of transcription , nuclear export , and degradation . A major challenge is to obtain a global , quantitative view of these processes . Here we measured the genome-wide nucleocytoplasmic dynamics of mRNA in Drosophila cells by metabolic labeling in combination with cellular fractionation . By mathematical modeling of these data we determined rates of transcription , export and cytoplasmic decay for 5420 genes . We characterized these kinetic rates and investigated links with mRNA features , RNA-binding proteins ( RBPs ) and chromatin states . We found prominent correlations between mRNA decay rate and transcript size , while nuclear export rates are linked to the size of the 3'UTR . Transcription , export and decay rates are each associated with distinct spectra of RBPs . Specific classes of genes , such as those encoding cytoplasmic ribosomal proteins , exhibit characteristic combinations of rate constants , suggesting modular control . Binding of splicing factors is associated with faster rates of export , and our data suggest coordinated regulation of nuclear export of specific functional classes of genes . Finally , correlations between rate constants suggest global coordination between the three processes . Our approach provides insights into the genome-wide nucleocytoplasmic kinetics of mRNA and should be generally applicable to other cell systems .
The production , nuclear export and degradation of mRNA are key steps in the control of cytoplasmic mRNA levels . Steady-state levels of transcripts in the cytoplasm are determined by the rates of these three processes . Hence , our understanding of gene regulatory systems requires quantitative knowledge of the relative contributions of each of these steps . Comparison of the kinetic rate constants for these steps across genes may provide insights into mechanisms of differential gene regulation . Recent technological advances have enabled genome-wide measurements of mRNA dynamics [1 , 2] and subcellular distribution [3] . In particular , the utilization of 4-thiouridine ( 4sU ) as a reagent to metabolically label newly synthesized RNA has provided the means to monitor RNA dynamics with a minimal perturbation [1] . Using this approach , various fundamental kinetic rates of mRNA , such as synthesis , splicing and decay have been quantified at genome-wide level in a number of cell types from different species [4–9] . Global quantification of RNA kinetic rates has revealed at least four major biological insights: 1 ) different classes of genes utilize distinct kinetic strategies to sustain/alter their expression levels; 2 ) transcription is the primary determinant of the steady-state level of RNA/protein , with contributions much higher than degradation rates; 3 ) motif analyses and experimental approaches have identified a range of RNA binding proteins that regulate RNA stability [10 , 11]; 4 ) the average rates of RNA decay differ dramatically between species . These studies of mRNA kinetics have taken the total cellular mRNA as a single entity to calculate the overall turnover rate , overlooking nucleocytoplasmic transportation , which is thought to be a key aspect of mRNA dynamics and has been shown to be regulated by a variety of evolutionarily conserved mechanisms [12 , 13] . Here we combined metabolic labeling of mRNA with cellular fractionation to systematically determine mRNA transcription , nuclear export and decay rates for thousands of genes . We developed a mathematical framework that infers nucleocytoplasmic kinetic rate constants from such labeling and fractionation time series data . We chose Drosophila Kc167 cells as a representative model for metazoan cells , because of their ease to perform experiment and the availability of a wealth of genome-wide information . We report kinetic rate constants for 5420 genes and determine the relative contributions of each of transcription , nuclear export and decay to overall cytoplasmic abundance . Moreover , we uncover links between the three kinetic steps and transcript features , interactions of specific RNA-binding proteins and specific gene classes .
To obtain genome-wide measurements of the nucleocytoplasmic dynamics of mRNA we followed a strategy as outlined in S1 Fig . Briefly , we performed a time series of metabolic labeling of RNA in Drosophila Kc167 cells using 4-thiouridine ( 4sU ) . We then isolated nuclear and cytoplasmic fractions from cells at each time point , and determined the relative abundance of “old” ( unlabeled ) mRNA for thousands of genes by high-throughput sequencing , as a function of time in both fractions . We then fed these measurements into a computational model that describes the process of sequential mRNA transcription , export and decay as a set of differential equations . Parameter fitting of the model to the measurements yielded kinetic rate constants for each of these three steps . Below we describe each step of the approach in more detail . To label newly synthesized RNA we used 4sU , which is known to have no major effects on gene expression in Drosophila [14] . We further tested the impact of 4sU on gene expression of Kc167 cells by genome-wide comparison of mRNA expression levels between cells treated for 480 minutes with 4sU . The overall gene expression profile was not much affected by 4sU labeling ( Spearman’s ρ = 0 . 98; P = 0 ) ( S2A Fig ) . A small set of 59 genes that were influenced by 4sU labeling were excluded from subsequent analysis ( S2A Fig ) . We then exposed cells to 4sU for 0 , 30 , 90 , 180 , 300 , and 450 minutes and subsequently fractionated the cells into nuclear and cytoplasmic portions by hypotonic lysis and centrifugation . In each sample we spiked in a fixed amount of total RNA from the yeast Saccharomyces cerevisiae for normalization purposes , analogous to a previously reported approach [15] . The sum of nuclear and cytoplasmic portions showed very good genome-wide consistency with the unfractionated total transcriptome that was independently measured , indicating that the loss of RNA during fractionation was generally low ( ρ = 0 . 89; P = 0; S2B Fig ) . We removed 107 genes for which this consistency did not hold up ( S2B Fig ) . We then purified pre-existing ( i . e . , unlabeled ) RNA by removal of newly synthesized RNA through sulfhydryl conjugation and biotin-streptavidin pull-down [1] . Finally , we isolated poly-adenylated mRNA from the unlabeled fractions and subjected it to high-throughput sequencing . Because it is known that 4sU labeling shows a bias for long genes , we corrected for such bias as described previously [8] . From the changes in mRNA abundance in the two fractions over time we then inferred kinetic rate constants ( see below ) . We conducted these experiments as two biological replicates , and the reproducibility of the detected mRNA levels was high for all time points ( S3A–S3D Fig ) . In bulk , the reads of both nuclear mRNA and cytoplasmic mRNA showed a continuous decrease over time relative to the yeast spike-in , reflecting the expected replacement of unlabeled mRNA by labeled mRNA ( S3E Fig ) . The amount of unlabeled mRNA eventually asymptotes to a plateau of 7 . 3±1 . 2% ( S3F Fig; Methods ) , which may reflect a pool of highly stable transcripts or incomplete removal of labeled mRNA . In order to check the purity of the nuclear and cytoplasmic fractions we monitored intron:exon ratios for a number of transcripts by quantitative reverse transcription polymerase chain reaction ( qRT-PCR ) . This revealed predominant presence of introns in the nuclear fraction , as expected ( S3G and S3H Fig ) . Furthermore , analysis of the high-throughput sequencing reads indicated a substantial enrichment ( 9 . 3±2 . 0 fold ) of rRNA in the cytoplasmic fraction ( S3I Fig ) . These results indicate that our measurements of pre-existing mRNA abundance over time in both the nuclear and the cytoplasmic compartments were generally robust . Subsequently , we applied mathematical modeling to the time series measurements to estimate rates of transcription , nuclear export and cytoplasmic decay for each transcript . We designed a set of first-order ordinary differential equations to describe the nucleocytoplasmic dynamics ( Fig 1A; see Methods ) . We assumed a steady-state model of a pool of non-synchronously dividing cells in which mature transcripts are produced in the nucleus , transported to the cytoplasm , and degraded in the cytoplasm , with each step described by a first-order reaction rate constant . In this model we assumed that transport of mRNA across the nuclear pore complex is unidirectional , which is generally supported by previous studies [12 , 13 , 16–19] . However , because transcripts redistribute between the nuclear and cytoplasmic compartments in the period between nuclear envelope breakdown and reformation during mitosis , we included kinetic terms to model this process ( Fig 1A ) . Furthermore , we followed the prevailing model that degradation of polyadenylated mRNA occurs predominantly in the cytoplasm [20 , 21] . Lastly , because it is not feasible to accurately quantify the abundance of every alternative transcript , we combined sequence reads from alternative transcripts , yielding a single kinetic model for each gene . An analogous mathematical framework based on similar assumptions was reported recently [22] . For each gene , we fitted this model to the experimental data from each of the two biological replicates ( S1 Table ) . As examples , we show the fitting results of the genes Arc1 and Bacc ( Fig 1B and 1C ) . Genome-wide , the goodness of fit was high , as indicated by the coefficients of determination ( r2 ) globally being close to 1 for both nuclear and cytoplasmic compartments ( Fig 1D and 1E ) . For subsequent analyses , out of 5 , 730 genes that were detected in all the time points of measurements after correction for 4sU labelling bias and exclusion for 4sU labeling and fractionation influence , we retained 5420 genes that have r2 > 0 . 8 . To test whether the data are compromised by undersampling , we down-sampled the sequencing reads to only 25 percent and repeated the modeling . The resulting estimated kinetic rates are generally consistent with the results based on the full dataset , demonstrating the robustness of the modeling ( S4A–S4C Fig ) . Analysis of the genes that did not fit the first order kinetic model well ( r2 < 0 . 8 ) indicated that they have substantially lower transcription rates ( S5A Fig ) . This could mean that the modeling is less accurate at low expression levels . However , genes with relatively poor fits also tend to have long introns ( S5B Fig ) ; we speculate that processing of these mRNAs is more complex and cannot be captured accurately by our computational model . Importantly , for the set of 5 , 420 genes that have r2 > 0 . 8 , all three modeled parameters have very good reproducibility between the two replicates ( Fig 1F–1H ) , ruling out overfitting as the cause for the good agreement between the modeled values and experimental data . We therefore used these 5 , 420 genes for subsequent biological analysis . Due to normalization to the spiked-in yeast mRNA , transcription is expressed in arbitrary units per minute . We emphasize that these transcription rates refer to the speed of production of polyadenylated mature mRNA , not the distance travelled by RNA polymerase as a function of time . The rates of export and cytoplasmic decay are expressed as fraction per minute , with a median value of 0 . 83% per minute ( corresponding to a half-life of 1 . 4 hours ) , and 1 . 40% per minute ( a half-life of 0 . 8 hours ) , respectively . For the majority of genes , the rate constants of transcription , export and cytoplasmic decay span about 5 , 0 . 5 and 0 . 9 orders of magnitude , respectively ( Fig 1F–1H ) . This prompted us to calculate the relative contributions of the three processes to the genome-wide variance in steady-state mRNA abundance . The results ( Fig 1I ) indicate that transcription explains most of the variance ( 89 . 4% ) , while the contribution of cytoplasmic decay is lower but still substantial ( 10 . 1% ) . In contrast , the contribution of nuclear export to the genome-wide variance is negligible ( 0 . 5% ) . Our estimation of the relative contribution of mRNA decay to the steady state mRNA abundance is lower than previously estimated for yeast ( ~30% ) [8] , but higher than estimated for mouse embryonic stem cells ( ~1 . 4% ) [23] . Next , we sought to identify potential determinants of individual kinetic steps . First , we investigated a possible relationship between transcript length and kinetic rates . This revealed that transcription rate has a considerable negative correlation with mRNA length ( ρ = -0 . 36; P = 1 . 1E-164; Fig 2A ) , suggesting that mature transcripts are generally less efficiently produced from long genes than from short genes . In part , this may be explained by a more extensive ( co-transcriptional ) splicing of long transcripts . Indeed , we find that transcription rate is negatively correlated with intron content ( ρ = -0 . 24; P = 1 . 1E-69; Fig 2B ) and with the number of exons ( ρ = -0 . 22; P = 1 . 8E-60; Fig 2C ) . This is in agreement with observations that elongation tends to slow down at exons [24] and that transcribed length has negative relationship with the rapidity of RNA polymerase II ( Pol II ) recruitment [25] . Export rates show no correlation with total mRNA length ( ρ = 0 . 01; P = 0 . 6 ) , suggesting that mRNA length is not a major limiting factor for transport through the NPC . However , export rate does show a notable negative correlation with the length of the 3’ UTR ( ρ = -0 . 26; P = 4 . 6E-85; Fig 2D ) and to a much lesser extent with the length of the coding region ( ρ = 0 . 13; P = 3 . 2E-22 ) or 5' UTR ( ρ = -0 . 12; P = 4 . 3E-18 ) . We speculate that the binding of regulatory proteins to the 3’ UTR may slow down export or actively retain transcripts in the nucleus . Interestingly , cytoplasmic decay rate shows a considerable positive correlation with total mRNA length ( ρ = 0 . 41; P = 3 . 8E-215; Fig 2E ) . We propose two possible explanations for this surprising link . First , cytoplasmic mRNA degradation may be initiated by stochastic attack by an endonuclease . In this model , long mRNAs simply have a higher probability to be cleaved than short mRNAs . The decay rate generally scales sub-linearly with mRNA length , as indicated by a linear regression slope of 0 . 25 in log-log space ( Fig 2E ) . This may reflect that mRNA is generally folded in a three-dimensional ribonucleoprotein particle , and the proportion of the mRNA that is buried inside this particle may increase with the linear length . A second , not mutually exclusive explanation may be that long mRNA molecules are more likely to contain motifs that have affinity to proteins that target the mRNA to the cytoplasmic decay machinery . Interestingly , decay rate has virtually no correlation with the length of the 3’UTR ( ρ = -0 . 05; P = 8 . 2 E-4; Fig 2F ) which is the primary site where miRNAs act [26 , 27] . miRNA-directed degradation may therefore not be the chief mechanism for cytoplasmic decay in Drosophila Kc167 cells . RNA-binding proteins ( RBPs ) are well known for their regulatory roles in specific steps of RNA metabolism [28–30] , but genome-wide assessment has not been yet carried out . We took advantage of recently published transcriptome-wide RNA interaction profiles of 20 RBPs [31] to uncover putative links with kinetic properties of mRNA . The interaction profiles were generated from Drosophila S2 cells , which are similar to the Kc167 cells that we used in our study . We compared the median kinetic rates of mRNAs that are bound and not bound by each RBP ( Fig 3A , 3C and 3D ) . This revealed that about two-thirds of the RBPs are significantly correlated with each step of kinetic regulation . The differences range from ~8 . 4 fold for transcription , to ~1 . 2 fold for export and ~1 . 9 fold for cytoplasmic decay . Overall , the correlations between RBP binding and the rate constants were similar for long and short transcripts ( S6A–S6F Fig ) , indicating that transcript length is not a substantial confounding factor in this RBP analysis . The RBP that is associated with the highest transcription rate is Cbp20 , a key component of the nuclear cap-binding complex ( Fig 3A ) . Its human homolog has previously been reported to facilitate Pol II release from promoters through interaction with transcription elongation through P-TEFb [32] , which should lead to reduction of the fraction of paused Pol II . We investigated this by computing the ‘pausing index’ of Pol II for each gene [33] . This pausing index is generally inversely correlated with transcription rates , and we found Cbp20-bound transcripts to have the lowest pausing index ( Fig 3B ) . It thus is likely that Cbp20 also has a role in releasing paused Pol II in Kc167 cells . The Exon Junction Complex protein Upf1 that links to nonsense-mediated decay also correlates with high transcription ( Fig 3A ) and a relatively low pausing index ( Fig 3B ) . A previous study in S . cerevisiae indicated that RNA decay factors can boost transcription independent of their function in RNA degradation [34] . The RBPs that are associated with the slowest transcription rate are msi and elav , which have little overlap in their mRNA binding specificities ( S6J Fig ) . However , these two proteins are expressed at extremely low levels in Kc167 cells [35] . It is therefore unlikely that they act as repressors of transcription . We speculate that their target mRNAs may in fact require binding of elav/msi for efficient transcription , and thus may be transcribed at higher levels in the cell types where these proteins are present ( e . g . , nervous system cells ) . We tested this hypothesis by comparing the expression level of transcripts in Kc167 that were shown to be bound by the ectopically expressed msi and elav [36] , to that of ML-DmBG2-c2 , a cell line representing cells of the central nervous system where these two proteins are expressed . Indeed , those genes that can be bound by msi and elav show statistically higher expression in ML-DmBG2-c2 cells ( msi: P = 1 . 9e-7; elav: P = 2 . 2e-4; two-sided Wilcoxon test ) . This suggest that msi and elav can promote transcription in the corresponding tissues , but further experimental evidence is needed . Remarkably , we found that all the 12 RBPs known to be involved in splicing [31] are associated with somewhat higher export rates . These RBPs , which tend to have overlapping mRNA binding specificities ( S6J Fig ) occupy the top 12 positions when ranked by mean export rate ( Fig 3C ) . This is in agreement with previous reports that have linked splicing factors to nuclear export [37 , 38] . In particular , among the splicing factors , SF2 has been shown as the adapter protein for TAP-dependent mRNA export [39] . Paradoxically , we find that mRNAs from intron-containing genes are generally not more rapidly exported than mRNAs from intron-less genes; there is even a slight opposite trend ( S6G Fig ) , which may be due to the fact that intronless genes generally have shorter 3’UTR ( S6H Fig ) , which in turn is associated with higher export rate ( Fig 2D ) . Analysis of the published RBP binding data [31] indicates that transcripts from genes with introns are not enriched for the binding of splicing-related RBPs , compared to transcripts from genes lacking introns ( S6I Fig ) . Together , these data suggest that splicing factors may promote export of mRNA at least in part independently of their role in splicing . The RBP that is associated with the highest cytoplasmic decay rate is CG6227 , a putative DEAD-box containing RNA helicase with so far unknown function ( Fig 3D ) . Lastly , the factor that is associated with the slowest cytoplasmic decay rate is Cbp20 . Transcripts with a 7-methylguanosine cap are thought to be bound by Cbp20 . This protein is generally restricted to the nucleus , while other proteins take over the cap-binding function in the cytoplasm [40] . Although it is not confirmed that this is also the case in Drosophila , it is therefore unlikely that Cbp20 directly affects cytoplasmic decay . Rather , Cbp20 binding to mRNA as detected in a total cell lysate [31] probably reflects the presence of 7-methylguanosine on the transcripts , and this capping is regulated in Drosophila cells and known to inhibit cytoplasmic decay [41] [42] . In conclusion , this analysis identifies candidate proteins that may control specific steps of nucleocytoplasmic mRNA kinetics . Chromatin is well-known for its role in regulating transcription , but there is also evidence that it may control the downstream fate of RNA . For example , specific histone modifications can influence alternative splicing [43 , 44] and promoter-bound proteins can direct cytoplasmic mRNA stability [45] [46] . We therefore asked whether the kinetic parameters derived from our measurements are correlated with the chromatin environments of the genes . To this end we stratified these parameters by the previously characterized five principal chromatin states [47] at both the transcription start sites ( TSSs ) and transcription termination sites ( TTSs ) . As expected , the modeled transcription rates differ widely among chromatin states , both at TSSs and TTSs ( Fig 4A–4D ) . BLUE chromatin , which is marked by H3K27me3 and Polycomb proteins , is associated with very low transcription rates . This is consistent with a wide body of literature indicating that Polycomb complexes directly repress transcription [48] . BLACK chromatin , a hitherto poorly characterized repressive chromatin type that carries H3K27me2 but not Polycomb proteins [14 , 47 , 49] , shows similarly low transcription rates , suggesting that BLACK chromatin also acts at the level of transcription . GREEN chromatin , marked by H3K9me2 and HP1 shows intermediate transcription rates , while the euchromatic YELLOW and RED states show high transcription activity , as expected . We also observed modest correlations between chromatin states and post-transcriptional kinetic parameters . For both TSS and TTS , transcripts arising from BLACK and BLUE chromatin showed lower export rates than those from YELLOW and GREEN chromatin ( Fig 4E–4H ) , but the difference is only ~1 . 1-fold . For cytoplasmic decay , we also observed significant chromatin-state-associated differences ( Fig 4I–4L ) , but only at TTSs and of minor magnitude ( up to ~1 . 2 fold ) . Notably , HP1-containing GREEN chromatin is associated with highest decay rates , which is consistent with a previous study that demonstrated that HP1 mediates heterochromatic transcript decay in S . pombe [50] . We obtained similar results with a 9-state chromatin state model , which is mostly based on histone modification maps [51] ( S7 Fig ) . In particular state 6 , roughly equivalent to BLUE chromatin , shows associations with low transcription and low export , especially when present at TSSs . Other states correlate with transcription levels but show only minor differences in export and decay rates . Together , these results indicate that chromatin states primarily affect transcription , and may have intriguing but subtle links to mRNA export and decay . Cross-talk has previously been observed between mRNA transcription and decay in yeast [3 , 9 , 34 , 52] . This prompted us to analyze possible relationships between the kinetic rate parameters in our data . Pairwise scatterplots revealed a number of interesting patterns . First , there is virtually no correlation between the rates of transcription and export ( ρ = 0 . 06 , P = 3 . 6E-05 , Fig 5A ) . Second , a moderate negative relationship exists between the rates of transcription and cytoplasmic decay ( ρ = -0 . 23 , P = 5 . 3E-66 , Fig 5B ) , which is in part due to a separate group of genes that we will discuss below . Third , we observe a positive relationship between rates of export and cytoplasmic decay ( ρ = 0 . 54 , P ≈ 0 , Fig 5C ) . These data suggest global coordination between mRNA decay and both transcription and nuclear export . The underlying mechanism is unclear; in yeast two subunits of RNA polymerase II have been implicated in such coordination [53–55] . One group of genes stands out in the scatterplots ( blue dots , Fig 5A–5C ) . Virtually all genes in this group encode for cytoplasmic ribosomal proteins ( cRPs ) . These genes are characterized by very high transcription , fairly low export , and very low cytoplasmic decay ( Fig 5A–5C ) . It is noteworthy that the cRP genes are completely separated from nuclear genes encoding mitochondrial ribosomal proteins ( mRPs ) , which also cluster in the scatterplots but show less extreme values ( green dots in Fig 5A–5C ) . The basis for this unique regulation is unclear , but we speculate that the 5'-terminal oligopyrimidine tract ( TOP ) motif , which is found in most cRP mRNAs [56 , 57] , plays a role in this . We searched for other functional categories of genes with distinct kinetic parameters by gene ontology ( GO ) analysis ( Fig 5D , S2 Table ) . Translational machinery genes , primarily comprising of cRP genes , exhibit relatively high transcription , low export and low cytoplasmic decay , as expected . Genes that are responsible for primary metabolic processes are highly transcribed , highly exported and slowly degraded , representing prominent efficiency of expression . The other house-keeping genes , including various synthetic and transport activities of macromolecules , are highly transcribed and lowly decayed but do not possess characteristic export rates . Lastly , it is intriguing that genes linked to neural differentiation and cellular response to stress are enriched for high export , and the latter process is also enriched for high decay . Presumably this provides routes to activate or inactivate gene regulatory cascades in a rapid and flexible manner when extrinsic stimuli are received in developmental processes or stress response . We note that genes with low transcription or high decay may be under-represented in this GO analysis , because they are less likely to have passed our stringent filters for model fitting . Overall , these results reveal curious links between specific functional gene modules and kinetic properties of their transcripts .
By combination of metabolic labeling , cell fractionation and mathematical modeling we determined key parameters of the nucleocytoplasmic dynamics of mRNA for 5 , 420 genes in Drosophila cells . Our subsequent analyses revealed that these kinetic rates are linked to various molecular components , have relationships with each other and are linked to specific biological processes . The export and cytoplasmic decay rates deduced from from our measurements and modeling are both on average in the range of ~1% per minute . These values are generally similar to rates estimated by previous studies . Genome-wide studies have determined median mRNA decay rates to range from 1 . 4% ( H . sapiens , [58] ) , 0 . 8% ( M . musculus , [6] ) , to 6 . 3% ( S . cerevesiae , [8] ) , while focused analyses of individual transcripts have yielded values ranging from ~1% per minute in zebrafish [59] to 1 . 67% per minute in mouse tissues [22] , which is very similar to our estimates . For nuclear export , recent microscopy studies estimated the retention time of mRNAs in human cells to be about 40–60 minutes , which corresponds to a rate of 1 . 2%-1 . 7% per minute [60] , and 8 . 6 minutes ( ~8% per minute ) in mouse tissue [22] . The latter export rate is somewhat higher than we typically observed , which may be explained by differences in cell type , species , or techniques used . Our estimates of transcription rates are in arbitrary units and can therefore not be compared to other studies . Our results indicate that nuclear export generally has a relatively minor impact on steady-state mRNA levels . Nevertheless , links with 3'UTR length and the binding patterns of RBPs point to mechanisms that regulate mRNA export . This is in line with previous gene-specific studies indicating that sequences in the 3'UTR of mRNA can affect nuclear export [61 , 62] . Our observation that several functional classes of genes show higher or lower export rates points to a certain degree of coordination of the export of mRNAs belonging to the same pathway . Some of the processes that we identified involve responses to DNA damage , stress and nutrients , as well as differentiation . This extends previous observations that the export of individual transcripts can be under control of such signaling events [62–67] . It will be interesting to study the changes in global nucleoplasmic kinetics of mRNAs when these pathways are activated by the appropriate stimuli . It is also noteworthy that transcripts derived from genes bound by Polycomb complexes ( BLUE chromatin ) show slightly slower export rates . We speculate that this may be caused by the broad affinity of Polycomb complexes for RNA [68 , 69] , which may lead to some sequestration of transcripts in the nucleus . This may reduce the availability of the transcripts in the cytoplasm to some degree . Another interesting possibility is that temporary nuclear retention of transcripts may buffer bursts of transcription [22] . We observed that the rate of cytoplasmic decay is positively correlated with transcript length . This is somewhat surprising , because it is generally thought that degradation of mRNA is primarily mediated by the 3' exonuclease activity of the exosome [8 , 70–72] , which is unlikely to lead to a faster decay for a longer RNA . Rather , the positive correlation with transcript length points to decay mediated by stochastic activity of an endonuclease [73] . A candidate for such endonuclease activity is the Drosophila exosome subunit Dis3 , which harbors ribo-endonuclease activity and is expressed in Kc167 cells [74] . In summary , we outlined a generally applicable experimental strategy and a mathematical framework to determine important parameters of nucleoplasmic dynamics for thousands of genes . One possible extension of our strategy is to quantify both the unlabeled and the 4sU-labeled mRNA fractions over time , rather than the unlabeled fractions alone . This may provide an even more precise view of the nucleoplasmic kinetics , particularly of transcripts with high transcription and export rates . The dataset reported here–as well as the uncovered links with mRNA characteristics , RBPs , and chromatin states–provides a foundation to begin to untangle the underlying mechanisms . Application of this approach to other cell types and species will help to understand the global principles of mRNA regulation in the context of differentiation and evolution .
Drosophila Kc 167 were cultured as previously described [75] . Around 1 million Drosophila kc167 cells were separately labelled in 5 ml medium in 10 cm culture dishes with 300 μM 4-thiouridine ( Sigma-Aldrich , Cat No . T4509 ) for 0 , 30 , 90 , 180 , 300 , 450 minutes . Cells were spun down , washed with serum-free medium and suspended with 120 μl hypotonic buffer consisting of 10mM NaCl , 2mM MgCl , 10mM Tris-HCL ( pH = 7 . 8 ) , 5mM dithiothreitol ( DTT ) , 0 . 5% nonylphenoxypolyethoxylethanol ( NP-40 ) . Suspensions were put on ice for 5 min and spun down at 2000 g at 4 degree for 5 minutes . Supernatants were taken out as cytoplasmic fraction and the pellets were suspended in 120μl hypotonic buffer as nuclear fraction . 700 μl TRIsure ( BIOLINE , Cat No . BIO-38032 ) containing 1 ng/μl total RNA from S . cerevisiae as spike-in was added to both fractions and RNA extractions were performed following the protocol of a published study [1] . 1 μg of nuclear and cytoplasmic RNA samples were reverse-transcribed ( BIOLINE , Tetro reverse transcriptase , Cat No . BIO-65050 ) with random-hexamers ( BIOLINE , Cat No . BIO-38028 ) . The reaction was subsequently diluted 20 times with water , 4 μl of which was used for qPCR . The primers used are: Lam exon forward , GAAGACCTGAATGAGGCGCT; Lam exon reverse , TGGTGTTCTCCAGGTCAACG; Lam intron forward , AAGTGCGTGGAAACTGAATCG; Lam intron reverse , CTTGCTTGAAACCACGCCTT; Fmo-2 exon forward , TGATGCAGTGCTTCCACAGT; Fmo-2 exon reverse , ATGTTCTGCACCGGCTACAA; Fmo-2 intron forward , GGCCCCGTGAGATCGATTAG; Fmo-2 intron reverse , TGGTAGCGACGTCACGTATT . Nuclear and cytoplasmic RNA were labeled with EZ-Link™ HPDP-Biotin ( ThermoFisher Cat No . 21341 ) and pre-existing RNA were purified by removal of biotinylated newly-synthesized RNA as previously described [1] . Total RNA was directly extracted from unfractionated cells following the protocol of a published study [1] . Polyadenylated RNA was purified by oligo-dT beads for both nuclear and cytoplasmic fractions , reverse transcribed ( SuperScript II Reverse Transcriptase , Invitrogen , # 18064–014 ) and constructed into strand-specific libraries using the TruSeq Stranded mRNA sample preparation kit ( Illumina Inc . , San Diego , RS-122-2101/2 ) according to the manufacturer's instructions ( Illumina , # 15031047 Rev . E ) . The generated cDNA fragments were 3' end adenylated and ligated to Illumina Paired-end sequencing adapters and subsequently amplified by 12 cycles of PCR . The libraries were analyzed on a 2100 Bioanalyzer using a 7500 chip ( Agilent , Santa Clara , CA ) , diluted and pooled equimolar into a 12-plex for each replicate and subjected to sequencing with 50 base single reads on a HiSeq2500 using V4 chemistry ( Illumina Inc . , San Diego ) . The two replicates were sequenced in two separate lanes . The total reads for the two replicates are 182 , 747 , 266 and 177 , 771 , 796 , with even reads distribution for each time point . Reads were mapped first to the transcriptome of S . cerevisiae ( Saccharomyces_cerevisiae . R64-1-1 . 78 ) and then to the transcriptome of D . melanogaster ( Drosophila_melanogaster . BDGP5 . 77 ) by Tophat [76] . For each time point , the number of reads that were mapped to Drosophila transcriptome was divided to the number of reads that were mapped to Saccharomyces transcriptome to obtain the factor for normalization . The number of Fragments Per Kilobase of transcript per Million mapped reads ( FPKM ) for each gene was calculated using the default setting of Cufflinks [77] and multiplied by the factor for normalization to obtain the relative abundance of transcripts . We intended to exclude genes that show strong anomalous behaviors which may be due to 4sU labeling or the fractionation procedure . Instead of linear exclusion , we also took the non-uniform distribution of dispersion into account . Considering two transcriptomes x and y of comparison , we used a simple non-linear function to depict the dispersion: f ( x ) =sd ( y ) =m+ne−px where x is the RNA abundance and f ( x ) is deviation from the perfect diagonal which corresponds to the standard deviation of y . The abundance of transcripts in the data were divided into 1000 intervals and the corresponding values were calculated . The non-negative parameters m , n , p were then fitted by the Levenberg—Marquardt algorithm using the ModFit function in the package of FME in R . We defined outliers as genes that exceed twice the amount of technical dispersion: outliers∣{y>x+2f ( x ) or x>y+2f ( y ) } The labelling bias of 4sU as function of the length of genes was corrected previously described [8] . The correction factor is calculated as F=1− ( 1−pr ) Nu where pr is the labeling probability that is determined to be around 0 . 01 [8] and Nu is the number of uridine in the transcripts of individual genes . To calculate the magnitude of remaining newly-synthesized RNA that contains 4sU after streptavidin removal , we assume first order turnover of total transcripts ( detailed in the next section of Quantitative modeling ) , and add a term for the remaining fraction of 4sU ( C ) that has not been removed by streptavidin pulldown , similar to a previous approach [6] . Using the notations from the next section of Quantitative modeling , the abundance of newly synthesized RNA over time is W4sU ( t ) =W0 ( egt−e− kTt ) Considering potential contamination factor ( U ) of W4sU into the unlabeled fraction , and assuming the pre-exsiting RNA follows first order turnover , the pre-existing fraction P is P ( t ) = ( W0−U⋅W4sU ( t ) ) e− kTt=W4sU ( t ) =W0 ( 1−Uegt+Ue− kTt ) e− kTt And the contamination fraction R ( t ) is R ( t ) = U⋅W4sU ( t ) =UW0 ( egt−e− kTt ) Therefore , the expected unlabeled fraction with contamination is W ( t ) =P ( t ) +R ( t ) The contamination factor U is estimated by minimizing d=1n∑i=1n ( L ( Wm ( ti ) logWm ( ti ) W ( ti ) ) 2 ) Where L is the loess function described in the Eq ( 18 ) of the next section and Wm ( ti ) is the measured abundance at a time point i . We used the Levenberg—Marquardt algorithm to fit experimental data using the ModFit function in the package of FME in R . U is estimated to be ( 7 . 3 ±1 . 2 ) % . The non-compartmentalized overall mRNA dynamics of non-synchronized Kc167 cells can be described by a simple ordinary differential equation dW ( t ) dt=kS−kTW ( t ) , ( 1 ) whereby for a given gene , W stands for the total amount of bulk mRNA with the unit of Fragments Per Kilobase Of Exon Per Million Fragments Mapped ( FPKM ) , t for time in minutes ( min ) , ks for transcription rate in FPKM·min-1 , kT for overall turnover rate in min-1 . The equation satisfies the quasi-steady-state assumption , i . e . , dW ( t ) dt=gW ( t ) , ( 2 ) since in standard culture medium , cells in the dish do nothing but merely doubling . Let g be the proliferation rate , which can be calculated from the measured doubling time of 24 hours of kc167 cells , g=100 ln224*60%=0 . 048%min−1 , ( 3 ) For pre-existing RNA , total amount Wp satisfies , dWp ( t ) dt=− kTWp ( t ) , ( 4 ) Denoting Wp ( 0 ) by W0 , Wp ( t ) =W0e− kTt . ( 5 ) Similarly , the simplest model for the nucleocytoplasmic dynamics of mRNA can be written as ( dN ( t ) dtdC ( t ) dt ) =[kS− ( kE+k′f ) ( kE+k′f ) kf− ( kD+kf ) ] ( N ( t ) C ( t ) ) =g ( N ( t ) C ( t ) ) , ( 6 ) whereby for a given gene , N and C stand for the total amount of mRNA in the , correspondingly , nuclear and cytoplasmic compartments in FPKM . For post-transcriptional kinetic rates , kE stands for exportation rate from the nucleus to the cytoplasm , kD for cytoplasmic decay rate , and kf stands for cytoplasm-to-nucleus inward transfer rate while kf′ for nucleus-to-cytoplasm outward transfer rate during mitosis , which will be discussed later . All post-transcriptional kinetic rates are in the unit of min-1 . The equation also satisfies the quasi-steady-state assumption , at t = 0 , denoting N ( t ) by N0 and C ( t ) by C0 , [kS− ( kE+k′f ) ( kE+k′f ) kf− ( kD+kf ) ] ( N0C0 ) =g ( N0C0 ) . ( 7 ) For pre-existing Np and Cp , ( dNp ( t ) dtdCp ( t ) dt ) =[− ( kE+k′f ) ( kE+k′f ) kf− ( kD+kf ) ] ( Np ( t ) Cp ( t ) ) . ( 8 ) Because Wp=Np+Cp , ( 9 ) rewrite Eq ( 8 ) in terms of Cp and Wp , ( dCp ( t ) dtdWp ( t ) dt ) = [− ( kE+k′f+kD+kf ) 0kE+k′f−kT] ( Cp ( t ) Wp ( t ) ) . ( 10 ) Therefore , dCp ( t ) dWp ( t ) = ( kE+k′f+kD+kf ) Cp ( t ) kTWp ( t ) −kE+ k′fkT , ( 11 ) from which we can get the analytical solution of Cp in terms of Wp Cp ( Wp ( t ) ) =C0 ( Wp ( t ) W0 ) kE+k′f+kD+kfkT+Wp ( t ) kE+k′f kE+k′f+kD+kf−kT ( 1− ( Wp ( t ) W0 ) kE+k′f+kD+kf−kTkT ) , ( 12 ) Because of Eq ( 5 ) , Cp ( t ) =C0e− ( kE+k′f+kD+kf ) t+W0 kE+k′fkE+k′f+kD+kf−kT ( e−kTt−e− ( kE+k′f+kD+kf ) t ) . ( 13 ) Similarly , Np ( t ) =N0e− ( kE+k′f+kf ) t+W0kfkE+k′f+kf−kT ( e−kTt−e− ( kE+k′f+kf ) t ) . ( 14 ) To determine the transfer rates of k′f , kf in mitosis , we considered the process of cell cycle . Because of the nature of the quasi-steady state in which stable proportionality of each phase of the cell cycle exists , we can determine during the doubling time of D = 24 hours , the duration of G1/S ( FG1/S ) takes about 20% of the time and G2/M ( FG2/M ) , in which cells have roughly two fold of cellular content compared to G1/S , takes about 80% of the time , based on published the FACS profile[78] [79] [80] . The duration of mitosis ( FM ) of drosophila cells takes around 1 hour . Kc167 cells have relatively large nuclei with the ratio of the diameters between the nucleus and the cell equals to rnc = 4:5 . Therefore , for every hour the cytoplasmic RNA that diffuses into the nucleus at the end of telophase is ( rnc3×2 ) D× ( FG1/S ×1+FG2/M ×2 ) ( N+C ) =2rnc3 ( N+C ) D ( FG1/S +2FG2/M ) . ( 15 ) And for every hour the nuclear RNA that diffuses into the cytoplasm at the beginning of M phase is 2ND× ( FG1/S ×1+ ( FG2/M −FM ) ×2 ) =2ND ( FG1/S +2FG2/M −2FM ) . ( 16 ) To obtain the kinetic rates of nucleocytoplasmic dynamics , we considered four attributes that ought to be satisfied , RNA-seq experiments render an over-dispersed non-Gaussian distribution for technical noise [81] . To adjust for this effect , we performed local polynomial regression fitting with the coefficient of variation ( CV ) with the mean value ( m ) for all the data points from the two biological replicates using the function loess in the package of stats in R , by which we generated function L that represents the numerical correspondence of loess . Thus , taking differential dispersions at individual time points into account we compute the difference on logarithmic scale , and minimized the corresponding four-component fitting gradient by least square . We used the Levenberg—Marquardt algorithm to fit experimental data using the ModFit function in the package of FME in R . To investigate the robustness of modeling in relation to the depth of sequencing , we randomly down-sampled the sequencing reads to only 25 percent of the original number . In this case , the number of genes that were detected in all samples after length correction was reduced to 3519 , of which 3403 pass the threshold of r2 > 0 . 8 . The consistency of the modeled rates was very high between the original and the down-sampled data , indicating that the performance of modeling is quite resilient to the reduction of sequencing depth ( S4A–S4C Fig ) . Annotations from BioMart ( http://www . biomart . org/ ) for Drosophila melanogaster genome BDGP5 were used for these analyses . For every gene , the length of transcript , intron , 3’UTR and 5’UTR and the number of exons were defined by the maximal values in each category from BioMart annotations . Spearman’s and Pearson’s correlations were calculated to associate length with kinetic rates . RBP binding data were retrieved from Stoiber et al [31] , in both binary form ( bound and unbound ) and quantitative form ( binding scores of the bound genes ) . To compare the kinetic difference between transcripts bound and unbound by a specific RBP , binary information was used to stratify the genes into two groups and two-sided Wilcoxon rank sum test were performed to calculate the statistical significance that was adjusted with the Holm—Bonferroni correction for multiple comparisons . Because RBP binding may have been underestimated for the 30% lowest expressed transcripts ( Figure S2 of [31] ) , we restricted this analysis to the 70% highest expressed genes ( 4457 genes total ) . Overlapping binding of RBP X and RBP Y were calculated as Overlap ( X , Y ) =No . of genes bound by X and YNo . of genes bound by X Overlap ( Y , X ) =No . of genes bound by X and YNo . of genes bound by Y To compute the pausing index of Pol II , we used Pol II Chip-seq data of Drosophila Kc167 cells from the modENCODE project ( DCCid: modENCODE_5569 ) , and calculated the ratio of Pol II signal within 200 bp around TSS and Pol II signal from 201bp to the end of the gene , similar to previous studies [33] . To investigate the relationship between binding strength of a specific RBP and kinetic rates , Spearman’s correlations were calculated . Chromatin states data were from Filion et al [47] and Kharchenko et al [51] . For every gene , the coordinates of the most 5’ TSS and the most 3’ TTS from BioMart annotation were defined as the TSS and TTS of the gene , for which corresponding chromatin states were assigned . kinetic rates associated with each chromatin state were compared by ANOVA and the statistical significance was calculated with Tukey’s range test . GO analysis was performed using the single ranked list method on the Gorilla server ( [82] , http://cbl-gorilla . cs . technion . ac . il/ ) . Corresponding p values were retrieved and gene ontology processes with p < 10−11 for at least one kinetic processes were displayed in a heatmap generated by the ‘pheatmap’ R package . | All mRNAs start from production in the nucleus , undergo exportation through nuclear pores and finally are degraded in the cytoplasm . A comprehensive characterization of the kinetic rates of all mRNAs is an important prerequisite for a global understanding of the regulation of the transcriptome and the cell . By conducting a time-series experiment and building a mathematical model , we trace the dynamics of mRNAs from the nucleus to the cytoplasm and determine the rates at each kinetic step at transcriptome-wide level . This information allows us to associate mRNA kinetic rates with a wealth of biological features and made some intriguing discoveries . We show mRNA decay is positively linked to transcript length while mRNA export is negatively linked to the length of the 3' UTR . We show binding of splicing factors is associated with faster rates of mRNA export . We provide evidence for global coordination between nuclear export an decay of mRNA . We show genes sharing specific functions tend to have similar nucleoplasmic kinetics , in which ribosomal proteins possessing special kinetic features exclusively stand out . Altogether , our integrated approach to quantitatively determine the rates of kinetic steps on a gene-by-gene basis provides a blueprint to obtain the global understanding of RNA regulation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
]
| [
"invertebrates",
"rna-binding",
"proteins",
"radiochemistry",
"messenger",
"rna",
"animals",
"invertebrate",
"genomics",
"dna",
"transcription",
"nuclear",
"decay",
"animal",
"models",
"drosophila",
"melanogaster",
"model",
"organisms",
"experimental",
"organism",
"systems",
"epigenetics",
"chromatin",
"drosophila",
"research",
"and",
"analysis",
"methods",
"genome",
"complexity",
"genomics",
"chromosome",
"biology",
"proteins",
"gene",
"expression",
"chemistry",
"insects",
"animal",
"genomics",
"arthropoda",
"physics",
"biochemistry",
"rna",
"cell",
"biology",
"nucleic",
"acids",
"nuclear",
"physics",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"biology",
"introns",
"organisms"
]
| 2017 | Comprehensive analysis of nucleocytoplasmic dynamics of mRNA in Drosophila cells |
Uterine leiomyosarcomas ( ULMSs ) are aggressive smooth muscle tumors associated with poor clinical outcome . Despite previous cytogenetic and molecular studies , their molecular background has remained elusive . To examine somatic variation in ULMS , we performed exome sequencing on 19 tumors . Altogether , 43 genes were mutated in at least two ULMSs . Most frequently mutated genes included tumor protein P53 ( TP53; 6/19; 33% ) , alpha thalassemia/mental retardation syndrome X-linked ( ATRX; 5/19; 26% ) , and mediator complex subunit 12 ( MED12; 4/19; 21% ) . Unlike ATRX mutations , both TP53 and MED12 alterations have repeatedly been associated with ULMSs . All the observed ATRX alterations were either nonsense or frameshift mutations . ATRX protein levels were reliably analyzed by immunohistochemistry in altogether 44 ULMSs , and the majority of tumors ( 23/44; 52% ) showed clearly reduced expression . Loss of ATRX expression has been associated with alternative lengthening of telomeres ( ALT ) , and thus the telomere length was analyzed with telomere-specific fluorescence in situ hybridization . The ALT phenotype was confirmed in all ULMSs showing diminished ATRX expression . Exome data also revealed one nonsense mutation in death-domain associated protein ( DAXX ) , another gene previously associated with ALT , and the tumor showed ALT positivity . In conclusion , exome sequencing revealed that TP53 , ATRX , and MED12 are frequently mutated in ULMSs . ALT phenotype was commonly seen in tumors , indicating that ATR inhibitors , which were recently suggested as possible new drugs for ATRX-deficient tumors , could provide a potential novel therapeutic option for ULMS .
Uterine leiomyosarcoma ( ULMS ) is a rare , highly malignant tumor that originates from the smooth muscle layer of the uterus , the myometrium . It is the most common subtype of uterine sarcoma and accounts for 1–2% of all uterine malignancies with an estimated incidence of 0 . 4/100 , 000 women per year [1 , 2] . The majority of ULMSs occur in women over 50 years of age typically causing symptoms such as abnormal vaginal bleeding , palpable pelvic mass , and abdominal pain . These symptoms greatly resemble those of common benign uterine leiomyoma , making early diagnosis of ULMS difficult . Surgical resection is the primary treatment option , while the use of adjuvant therapies varies widely . ULMS show low sensitivity to both chemotherapy and radiation therapy [3 , 4] . In most cases , the diagnosis is made histologically after the surgery , and even then , the clinical course of ULMS is difficult to predict . Currently , the most prominent prognostic factors include stage , age , and tumor size [5–7] . The 5-year overall survival has remained <50% due to a high recurrence rate ( 53–71% ) and metastatic capacity [6 , 8] . Most ULMSs are aneuploid with both complex numerical and structural chromosomal aberrations [9] . Albeit no consistent structural aberrations have been identified , abnormalities affecting chromosomal regions 1p , 10q , 13q , and 14q have been observed in multiple cases [10] . So far , only a few genes have been associated with this tumor type , including tumor protein P53 ( TP53 ) , RB1 , MDM2 , CDKN2A , and KIT [9 , 11] . These are all common cancer genes not specific for smooth muscle malignancies and the exact molecular mechanisms underlying ULMS tumorigenesis remain elusive . During the last decade , next-generation sequencing technologies have increasingly provided genome-wide data on somatic landscapes in various cancer types enabling the discovery of novel cancer genes and mechanisms with important prognostic and therapeutic implications [12] . Here , we performed exome sequencing on 19 ULMSs to further elucidate the molecular etiology of these tumors , identifying frequent mutations in TP53 , alpha thalassemia/mental retardation syndrome X-linked ( ATRX ) , and mediator complex subunit 12 ( MED12 ) . This is the first description of high-throughput sequencing on ULMSs .
We performed exome sequencing on genomic DNA of 19 formalin-fixed paraffin-embedded ( FFPE ) ULMSs . The average coverage of captured exonic regions reached a mean depth of 21x and 92% of the captured regions had a minimum coverage of four reads ( S1 Table ) . After filtering the exome sequencing data , we observed a mean of 373 somatic mutations per tumor ( range 240–779 ) . The majority of mutations in each tumor specimen represented single-nucleotide variations ( ∼88%; range 81–95% ) , while deletions accounted for ∼9% ( range 4–15% ) and insertions ∼3% ( range 1–7% ) ( S1 Table ) . Two tumors , LMS49 and LMS51 , harbored more mutations than other ULMSs , but the mutation spectrum followed the common pattern . In the exome sequencing data analysis , we focused on genes that were mutated in at least two tumors . This resulted in a list of 43 genes ( S2 Table ) . The majority of these genes ( 37/43; 86% ) were mutated in two tumors , while six genes , TP53 , ATRX , MED12 , fibrous sheath interacting protein 2 ( FSIP2 ) , ATP-binding cassette , sub-family A ( ABC1 ) , member 13 ( ABCA13 ) , and ankyrin repeat domain 26 ( ANKRD26 ) , had mutations in three or more tumors ( Fig 1 ) . The most frequently mutated gene was TP53 , which was mutated in six tumors ( 6/19; 32% ) ( S1 Fig ) . Two mutations were nonsense mutations creating a premature stop-codon and four were missense alterations; all missense changes were predicted pathogenic by two independent in silico tools ( S2 Table ) . All the observed TP53 mutations have been reported as somatic mutations in the COSMIC-database . The second most commonly mutated gene was ATRX , which was mutated in five tumors ( 5/19; 26% ) ( S1 Fig ) . The total number of mutations was six as one tumor ( LMS71 ) contained two distinct mutations . All mutations were either nonsense mutations or small frameshift insertions or deletions , and were thus predicted to result in a truncated protein product . As ATRX mutations have been associated with alternative lengthening of telomeres ( ALT ) , we specifically searched the exome sequencing data for possible mutations in death-domain associated protein ( DAXX ) , as also these mutations have been associated with the ALT phenotype [13 , 14] . Indeed , one ULMS ( LMS61 ) had a mutation in DAXX . This mutation was a nonsense mutation ( Glu650Stop ) most likely leading to a truncated or unstable protein product . Four mutations ( 4/19; 21% ) were observed in MED12 ( S1 Fig ) . All these were missense changes affecting amino acids Gly44 ( 3 mutations ) or Leu36 ( 1 mutation ) , which have previously been reported as mutational hotspots in uterine leiomyomas [15] . All mutations were predicted to have a deleterious effect on protein function ( S2 Table ) . Neither MED12 , TP53 , nor ATRX mutations were mutually exclusive ( Fig 1 ) . Alterations in FSIP2 ( 4/19; 21% ) , ABCA13 ( 3/19; 16% ) , and ANKRD26 ( 3/19; 16% ) all represented missense changes that scattered along the gene lengths . Two tumors had the same Met487Ile substitution in ANKRD26 . One alteration ( Gln581Leu ) in FSIP2 and all changes in ABCA13 were predicted pathogenic by both Polyphen-2 and SIFT , whereas none of the other variants were predicted damaging by both in silico tools . We evaluated the protein expression levels of TP53 , ATRX , and DAXX in the 19 exome-sequenced ULMSs by immunohistochemistry and validated the results in a larger set of 33 additional tumors ( S2 Fig and S3 Table ) . DAXX immunostaining was successful in all 52 tumors and interpretable results for TP53 and ATRX were obtained from 50 and 44 tumors ( 50/52 , 96%; 44/52 , 85% ) . Aberrant TP53 expression was observed in 33 out of 50 ULMSs ( 66% ) ( S3 Table ) . Twenty-three out of 44 successfully analyzed ULMSs ( 52% ) showed loss of nuclear ATRX expression , including all immunohistochemically successful ATRX mutation-positive tumors . Clearly diminished DAXX expression was present in only one ULMS ( 1/52 , 2% ) ( Fig 2A and 2C ) : a tumor with the nonsense mutation ( S3 Table ) . Telomere-specific fluorescence in situ hybridization ( FISH ) was carried out to assess the potential effect of ATRX and DAXX mutations on telomere length ( Fig 2B and 2D ) . Twelve out of 19 exome-sequenced ULMSs ( 63% ) were ALT-positive ( S3 Table ) . This included four out of five ATRX mutation-positive tumors ( 80% ) as well as the one DAXX mutation-positive tumor . Also seven out of 13 cases ( 54% ) without detectable ATRX or DAXX mutations showed ALT positivity . Loss of ATRX or DAXX expression seems to correlate very well with the ALT phenotype . Kaplan-Meier survival curves were generated to study the association between TP53 and ATRX expression and overall survival time . The median overall survival time for all patients was 65 months ( 95% confidence interval 31 . 7–98 . 3 months ) . Only the number of Stage I tumors was large enough for the analyses . Neither aberrant TP53 or ATRX expression associated with poor survival ( P = 0 . 261 , P = 0 . 127 ) ( Fig 3 ) . Of note , TP53 and ATRX expression statuses correlated with each other ( P = 0 . 005 ) .
In this study , we examined somatic variation in 19 ULMSs by exome sequencing . We focused on genes , which were mutated in at least two tumors; altogether 43 such genes were identified . The most frequently mutated genes included TP53 , ATRX , MED12 , FSIP2 , ABCA13 , and ANKRD26 . TP53 was the most commonly mutated gene with 32% of the tumors harboring mutations . Alterations in TP53 have been previously implicated in leiomyosarcomas and suggested to play a role in leiomyosarcoma pathogenesis [16–18] . In this study , most mutations ( 67% ) were missense changes located in exons 4–8 . This is in line with previous studies , where the majority of mutations have been missense mutations in exons 5–8 , the most highly conserved region of the gene [9 , 17 , 19] . These mutations are known to alter the protein structure , inhibit its tumor suppressor function , and result in its prolonged half-life . Immunohistochemical analysis including 50 ULMSs confirmed altered expression in the majority of tumors , highlighting the role of TP53 in ULMS development . ATRX was the second most frequently mutated gene with mutations observed in five tumors ( 26% ) . All mutations were either nonsense or frameshift alterations most likely leading to a truncated protein product . Loss of ATRX expression has been reported in leiomyosarcomas of various sites [20–22] and a recent meeting abstract on ULMSs reported genomic alterations of this gene in 32% ( 8/25 ) of the studied tumors , supporting our findings [23] . We successfully analyzed ATRX protein levels in 44 ULMSs and showed that 52% of the tumors , including all reliably analyzed mutation-positive lesions , had clearly reduced expression . ATRX encodes a transcriptional regulator that contains an ATPase/helicase domain , and is thus a member of the SWI/SNF family of chromatin remodelling proteins . Loss of ATRX expression has been associated with ALT [13 , 24] , which prompted us to analyze the telomeres with telomere-specific FISH . The ALT phenotype was confirmed in all ULMSs with diminished ATRX expression . Some exome-sequenced tumors with reduced ATRX expression and ALT positivity did not show ATRX mutations , suggesting that there are regulatory or larger structural alterations undetectable by exome sequencing , or that the quality of FFPE samples was inadequate to reveal the underlying mutation . Interestingly , the only ULMS with two ATRX mutations did not show ALT . ATRX is known to functionally cooperate with DAXX and DAXX mutations have been associated with ALT [13 , 14] . We therefore scrutinized the exome data for possible DAXX mutations . One nonsense mutation was identified , and FISH confirmed the ALT phenotype . Overall , these results show that ALT is very common in ULMS and that in addition to ATRX , also DAXX mutations contribute to the phenotype . Importantly , ALT was recently suggested to render cancer cells hypersensitive to ATR inhibitors [25] . These inhibitors might provide a novel treatment for ULMS , in which chemotherapeutic options have thus far been limited . MED12 was mutated in four ULMSs ( 21% ) . All mutations were in exon 2 , which is a known mutational hotspot in MED12 . These mutations were first observed in uterine leiomyomas [15] , and subsequently they have been identified in other tumor types [26] . Previous screening studies have reported recurrent MED12 mutations also in ULMS with similar frequencies as observed here [26] . It may be that a subset of ULMSs arises through a leiomyoma precursor , or alternatively MED12 mutations may provide growth advantage to ULMSs . MED12 is part of a multi-protein complex Mediator , which plays a key role in global transcription regulation in eukaryotic cells [27] . Based on our results , MED12 mutations can co-occur with TP53 and ATRX mutations . FSIP2 , ABCA13 , and ANKRD26 were mutated in at least three tumors and additional 37 genes had mutations in two tumors . Most alterations were missense changes and gave either neutral or controversial results in in silico predictions . The possible role of these genes in ULMS development cannot be directly assessed as in addition of providing growth advantage to the cell , the observed alterations may represent rare germline polymorphisms or passenger mutations with no functional significance . Although aberrant expression of both TP53 and ATRX in Stage I ULMSs , the only group of tumors large enough for the analyses , did not associate with poor overall survival , a trend toward poorer survival was seen in the patients . The limited number of samples in the survival analyses and the observation that expression statuses were associated with each other makes it difficult to draw conclusions regarding prognostic implications of TP53 or ATRX expression levels . In general , TP53 alterations are the most common genetic changes in human cancers and they are particularly associated with an aggressive phenotype . Recently , loss of ATRX expression was associated with poor clinical outcome in ULMS [21 , 22] . Larger sample series with information on both TP53 and ATRX are required to confirm these findings . ULMSs are rare and aggressive cancers . In most cases the diagnosis is made only at surgery , and many patients thus present with an advanced disease . Here , we have utilized exome sequencing and identified several recurrently mutated genes , including TP53 , ATRX , and MED12 . While MED12 mutations are the most common alterations in benign conventional leiomyomas , TP53 or ATRX mutations have not been observed in these tumors . Specifically , identification of inactivating ATRX mutations and their association with the ALT phenotype in the substantial proportion of tumors may be translatable into clinical practice should the suggested effect of ATR inhibitors prove effective .
This study was approved by the appropriate ethics review board of Hospital District of Helsinki and Uusimaa , Finland ( 408/13/03/03/2009 ) . Fifty-two archival FFPE ULMS tissue samples were derived from the Department of Pathology , Hospital District of Helsinki and Uusimaa , Finland , according to Finnish laws and regulations by permission of the director of the respective health care unit . These specimens represented diagnostic ULMS samples collected during surgery in 1985–2013 . Simultaneously with the sample collection , clinical data were obtained for these cases ( Table 1 ) after which the samples were anonymized for the study . Nineteen ULMSs ( diagnosis 2003–2013 ) entered exome sequencing , while the remaining 33 tumors were available on a tissue microarray for immunohistochemistry . Before exome sequencing , hematoxylin-eosin-stained sections from each specimen were re-evaluated by a pathologist ( RB ) and verified as ULMSs according to the WHO criteria [28] . Tumor percentage was ≥90% in all samples . Genomic DNA was extracted with a standard phenol-chloroform method . Sample libraries were prepared using NEBNext DNA Library Prep Reagent Set for Illumina ( New England Biolabs Ltd . catalog# E6000 ) and subjected to exome capture with NimbleGen SeqCap EZ System ( Roche NimbleGen ) . Paired-end short read sequencing was performed with HiSeq 2000 ( Illumina Inc . ) at Karolinska Institutet , Sweden . Read mapping and somatic variant calling were carried out as previously described [29] . Additionally , single duplicate reads were removed with an in-house script . Exome data was analyzed with an in-house analysis and visualization tool RikuRator . The requirements to call a variant included a minimum coverage of six reads and the mutated allele to be present in at least 20% of the reads . The Genome Analysis Toolkit ( GATK ) quality score of variants was required to be 25 or above . Both exonic regions and sequences within three base pairs of the exon-intron boundaries were included in the study . Synonymous changes , variants present in the dbSNP database ( release 138 ) , and variants in Exome Aggregation Consortium Server with a frequency over 0 . 1% were disregarded . To remove other potential germline variants , the exome data was filtered against data from 2315 Finnish controls ( 93 individuals from the 1000 Genomes Project , 1941 individuals from The Sequencing Initiative Suomi ( SISu ) ( http://www . sisu . fimm . fi ) , and 281 in-house control exomes or genomes ) . Recently , it has been estimated that about 400 control samples remove germline variation ( single-nucleotide variants and indels ) from a tumor sample at least as efficiently as the matched normal sample [30] . Lastly , all the remaining variants were individually visualized with Rikurator to exclude those only present in the same direction reads as likely artifacts . The functional effects of the variants were predicted by two independent in silico tools: SIFT ( http://sift . jcvi . org/ ) and Polyphen-2 ( http://genetics . bwh . harvard . edu/pph2/ ) . All candidate variants in genes mutated in at least three tumors were validated by direct sequencing . Oligonucleotide primers were designed with Primer3Plus software ( http://www . bioinformatics . nl/cgi-bin/primer3plus/primer3plus . cgi ) ( S4 Table ) . PCR products were sequenced directly utilizing Big Dye Terminator v . 3 . 1 sequencing chemistry ( Applied Biosystems ) on an ABI3730 Automatic DNA Sequencer . TP53 , ATRX , and DAXX immunolabeling was performed on FFPE sections of all 19 exome-sequenced ULMSs and on a tissue microarray containing 33 additional tumors . For TP53 , immunostaining was performed as previously described [31] . For ATRX and DAXX , heat-induced antigen retrieval was carried out in a microwave using citrate buffer ( pH 6 . 0 ) for 20 min . Endogenous peroxidase blocking was followed by overnight incubation with the primary antibody at 4°C ( anti-ATRX 1:500 dilution , Sigma-Aldrich catalog# HPA001906; anti-DAXX 1:500 dilution , Sigma-Aldrich catalog# HPA008736 ) . The primary antibody was detected with DAB Plus Substrate System ( Thermo Fisher Scientific catalog# TA-060-HDX ) . Immunohistochemical scoring was assessed by a pathologist ( RB ) . Only nuclear labeling of the proteins was evaluated . The loss of nuclear staining in tumor cells together with retained expression in non-neoplastic cells ( endothelial or inflammatory cells ) was considered loss of expression . The scoring was done without knowledge of the clinical outcome data . Detection of large , abnormally intense , intra-nuclear telomere DNA aggregates via telomere-specific FISH is considered the most sensitive and specific marker for identifying ALT phenotype in fixed tissue samples [13] . FFPE sections were deparaffinized at room temperature with xylene ( 3x10 min ) and 100% EtOH ( 2x10 min ) and air-dried . Subsequently , the slides were rinsed in phosphate-buffered saline ( PBS ) at 37°C ( 2x5 min ) followed by RNAse A treatment ( Sigma-Aldrich catalog# R4642 ) at 37°C for an hour . After a series of washes at room temperature with saline-sodium citrate ( pH 7 . 0; 3x5 min ) and deionized water ( 2x5 min ) , the slides were digested with Digest All 3-pepsin ( Invitrogen/Life Technologies catalog# 00–3009 ) at 37°C for 10 min and rinsed with PBS at room temperature ( 2x5 min ) . Next , the slides were dehydrated and hybridized with Cy3-labeled peptide nucleic acid ( PNA ) probe ( Panagene Inc . catalog# F1006-5 ) . The denaturation took place at 85°C for 10 min and hybridization in dark at room temperature for an hour . Post-hybridization washes with saline-sodium citrate/0 . 1% Tween-20 ( pH 7 . 0; 2x10 min ) at 55°C and at room temperature for 10 min were followed by nuclear counterstaining with DAPI . The slides were imaged with a Zeiss Axio Imager epifluorescence microscope and image acquisition took place through Hamamatsu Orca Flash 4 . 0 LT camera and Zen software . The assessment of FISH slides was carried out independently by two authors ( NM , MA ) . ULMSs were classified as ALT-positive if ≥5% of 300 assessed neoplastic cells displayed ALT-associated , abnormally bright telomeric DNA aggregates . In all cases , regions of necrosis and overlapping cells difficult to interpret were excluded from consideration . Statistical analyses were performed using SPSS statistical software for Windows version 22 . 0 ( SPSS Inc . ) . Here , survival was defined as overall survival time from the time of diagnosis . Survival curves related to TP53 and ATRX expression were generated using the Kaplan–Meier method , and median survival times with 95% confidence intervals were given . Comparison of survival curves between normal and aberrant expression was performed using the log-rank test . P-value <0 . 05 was considered statistically significant . Association between TP53 and ATRX expression statuses was evaluated using cross tabulation and Fisher’s exact test . | Uterine leiomyosarcomas are rare , malignant smooth muscle tumors with a poor 5-year survival and high recurrence rate . They account for 1–2% of all uterine malignancies with an estimated incidence of 0 . 4/100 , 000 women per year . The symptoms and signs of this tumor type widely overlap with those of common benign uterine leiomyomas , making early diagnosis of uterine leiomyosarcomas difficult . Currently , the diagnosis of these tumors is often incidental and postoperative . Despite previous cytogenetic and molecular studies , their molecular background has remained elusive . Identification of novel molecular genetic characteristics in uterine leiomyosarcomas is clinically relevant to further improve the diagnosis and prognosis of the patients . Here , we performed exome sequencing on 19 tumors , revealing frequent mutations in TP53 , ATRX , and MED12 . The discovery of frequent inactivating ATRX mutations provides a potential novel therapeutic target for uterine leiomyosarcomas . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"cancer",
"detection",
"and",
"diagnosis",
"medicine",
"and",
"health",
"sciences",
"chromosome",
"structure",
"and",
"function",
"insertion",
"mutation",
"cancers",
"and",
"neoplasms",
"oncology",
"telomeres",
"mutation",
"nonsense",
"mutation",
"frameshift",
"mutation",
"chromosome",
"biology",
"leiomyosarcoma",
"sarcomas",
"cell",
"biology",
"phenotypes",
"genetics",
"biology",
"and",
"life",
"sciences",
"chromosomes"
]
| 2016 | Exome Sequencing of Uterine Leiomyosarcomas Identifies Frequent Mutations in TP53, ATRX, and MED12 |
Translational control and messenger RNA ( mRNA ) decay represent important control points in the regulation of gene expression . In yeast , the major pathway for mRNA decay is initiated by deadenylation followed by decapping and 5′–3′ exonucleolytic digestion of the mRNA . Proteins that activate decapping , such as the DEAD-box RNA helicase Dhh1 , have been postulated to function by limiting translation initiation , thereby promoting a ribosome-free mRNA that is targeted for decapping . In contrast to this model , we show here that Dhh1 represses translation in vivo at a step subsequent to initiation . First , we establish that Dhh1 represses translation independent of initiation factors eIF4E and eIF3b . Second , we show association of Dhh1 on an mRNA leads to the accumulation of ribosomes on the transcript . Third , we demonstrate that endogenous Dhh1 accompanies slowly translocating polyribosomes . Lastly , Dhh1 activates decapping in response to impaired ribosome elongation . Together , these findings suggest that changes in ribosome transit rate represent a key event in the decapping and turnover of mRNA .
Messenger RNA ( mRNA ) is targeted for destruction in a precise and regulated fashion . In eukaryotic cells , the digestion of the 3′ polyadenosine tail ( deadenylation ) is the first step , followed predominantly by removal of the mRNA cap and 5′→3′ exonucleolytic digestion or , rarely , 3′→5′ degradation catalyzed by the cytoplasmic exosome [1] . Decapping of mRNA , therefore , represents an important regulatory node in mRNA turnover and is , in most cases , both rate limiting and non-reversible [2] . In yeast , mRNA decapping is catalyzed by a single polypeptide encoded by DCP2 . DCP2 is conserved from yeast to humans , however it is becoming apparent that additional decapping activities exist in metazoans [3] . The rate at which an mRNA 5′ cap is removed is highly variable , and although not completely understood , the rate of Dcp2-dependent mRNA decapping is modulated by a suite of protein factors that facilitate the binding and catalytic activity of the decapping enzyme itself . Moreover , mRNA translation is critical in determining the overall level of decapping and stability of the mRNA [2] . mRNAs that initiate translation poorly are generally unstable and vice versa . The exact nature of the relationship between mRNA translation and decay is unclear , however it has been postulated that decapping activators may also function to promote mRNA turnover by monitoring mRNA translational status and/or promoting translation states that favor the decapping reaction . Of the many factors that influence mRNA decapping rates , the function of the DEAD-box RNA helicase Dhh1 most clearly ties mRNA decapping to protein synthesis . Dhh1 was first shown to be involved in modulating mRNA decapping in yeast [4] , [5] . At the same time , it was determined that Dhh1 homologues function as translational repressors in a variety of biological contexts . For example , the Xenopus ortholog of Dhh1 , Xp54 , was identified as a component of translationally silenced messenger ribonucleoprotein complexes ( mRNPs ) in Xenopus oocytes [6] . Moreover , the orthologous Drosophila protein , Me31b , is required for translational silencing of oskar mRNA and is , therefore , a critical determinant in defining the posterior pole in the fly embryo during development [7] . Subsequent studies indicated that Me31b also represents an important neurological factor through its regulation of CaMKII mRNA translation and association with the translational repressor , Fragile X Mental Retardation Protein ( FMRP ) [8] , [9] . Furthermore , depletion of the human Dhh1 ortholog , RCK/p54 [10] , or Xenopus Xp54 [11] leads to general derepression of mRNA translation . Finally , the role of yeast Dhh1 in promoting mRNA decapping was suggested to result from its role as a general translational repressor [12] . Together , these data demonstrate that Dhh1 and its homologues are a conserved family of translation regulatory proteins whose activity can lead to storage and/or destruction of translationally repressed mRNAs . Despite the widespread control on mRNA translation and turnover by Dhh1 proteins , the molecular mechanism by which it controls mRNA metabolism remains unclear . Several pieces of evidence have supported a model that Dhh1 proteins alter the association of translation initiation complexes with mRNA , thereby rendering the cap accessible to the decapping machinery [12] . Consistent with this , a direct competition exists between the mRNA decapping and translation initiation machineries for the mRNA cap [13] , [14] . Specifically , mRNA decapping rate is enhanced in vivo when translation initiation is impaired either in cis or trans . Moreover , the major cytoplasmic cap binding protein , eIF4E , competes with Dcp2 for association with the 5′ cap in vitro [15] . Thus , it has been proposed that association between translation initiation complexes and the mRNA must be antagonized before decapping can occur , a function that could be served by Dhh1 . Two studies have provided evidence that Xenopus Xp54 complexes with the eIF4E inhibitor , eIF4E-T , thereby providing a possible model for how Dhh1 proteins could block eIF4E function [16] , . In addition , experiments tethering Xp54 to an mRNA lead to the translational repression of capped mRNAs but not mRNAs lacking a 5′ cap or undergoing translation initiation using an internal ribosome entry site ( IRES ) element [18] . Lastly , recombinant Dhh1 inhibits 48S initiation complex formation in vitro [12] . The observation that decapping activators , including Dhh1 , Pat1 , and Lsm1 , can be found in cytoplasmic aggregates called Processing bodies ( P-bodies ) has also provided support for a model in which translation initiation is blocked prior to mRNA decapping [19] , [20] . P-bodies are proposed sites of mRNA decapping and degradation and encompass the full complement of decapping factors but are thought to be void of translation initiation factors and ribosomes [20] . In combination with the above work , this has led to a two-step model for mRNA decay in which deadenylation leads to the dissociation of mRNA from the translational apparatus and reorganization into a P-body where it is either stored or decapped and destroyed [20] . Importantly , the dissociation of ribosomes from the mRNA and mRNP remodeling have been hypothesized to be dependent on Dhh1 proteins [12] . Recent findings from a number of labs has , however , called into question the requirement for P-bodies in the translational repression and/or decay of mRNA , as these processes can be uncoupled from the accumulation of P-bodies in yeast and metazoans [21]–[23] . Under a common assumption that translation initiation is rate limiting for protein synthesis , repression of translation initiation prior to mRNA decapping would be predicted to result in ribosome run-off and decapping would occur predominately on ribosome-free mRNAs . In contrast , however , we have recently demonstrated that the majority of mRNA decapping occurs while mRNA maintains an association with polyribosomes , demonstrating that dissociation of mRNAs from ribosomes is not a prerequisite or general occurrence for mRNA decapping to occur [24] , [25] . Based on this and additional evidence , we evaluated a role for Dhh1 in mediating a translational repression event that does not promote the loss of ribosome and mRNA association . Here we show that Dhh1 functions in vivo primarily to repress mRNA translation and that its influence on decapping rate is predominantly a secondary effect . We demonstrate that Dhh1 inhibits mRNA translation in a manner independent of the translation initiation factors eIF4E and eIF3b . Consistent with the observation that mRNA decapping occurs on polyribosomes , tethering Dhh1 to an mRNA results in the accumulation of ribosomes on the mRNA . Moreover , endogenous Dhh1 protein associates with slowly moving polyribosomes . These data suggest that Dhh1 mediates a slowing of ribosome movement that may be a necessary first step before mRNA decapping can occur . Consistent with this , we show that slowing ribosome elongation in cis stimulates mRNA decapping in a Dhh1-dependent manner . Together , these data support a model that decapping of mRNA occurs on polyribosomes that have been impaired in ribosome transit in part by the activity of the general translational repressor Dhh1 .
It has been extensively documented that Dhh1 and its orthologs are integral components of the decapping complex [5] , [26] . Moreover , it has been observed that the homologs function as general repressors of mRNA translation [10]–[12] . The precise role for Dhh1 in this process has , however , remained elusive but has been suggested to involve remodeling of translation initiation factors at a step before 48S translation initiation complex formation on the mRNA [12] , [16]–[18] . Due to the competition that exists for the mRNA 5′ cap between translation initiation factors and the decapping machinery , remodeling of the mRNP at the cap may be sufficient to explain the bipartite role Dhh1 appears to play in promoting both mRNA decapping and translational repression [12] . We wished to experimentally separate the two known functions of Dhh1 to evaluate the mechanism by which Dhh1 mediates translational repression and/or mRNA decay on an individual mRNA . Since little is understood about recruitment of Dhh1 to mRNA , we utilized a tethered-function approach to directly assay the functional consequences of Dhh1 binding to a reporter mRNA independent of its natural recruitment [27] . This assay has successfully been used to dissect the role of numerous RNA binding proteins in a variety of biological contexts [28]–[30] . The bacteriophage MS2 coat protein alone ( MS2 ) or a protein chimera of Dhh1 and MS2 ( Dhh1-MS2 ) were expressed from plasmid vectors along with reporter mRNA harboring MS2 RNA recognition elements in its 3′ UTR . Three different reporter mRNAs were used in various assays ( Figure 1A ) . The first , MFA2 , expresses the unstable MFA2 mRNA with 3′ UTR MS2 binding sites [28] . The second and third represent MFA2 and PGK1 genes with their protein coding regions replaced by that of green fluorescence protein ( GFP; Figure 1A; M/GFP and P/GFP , respectively ) . This combination of reporters allowed measurement of the consequence of tethering Dhh1 on both mRNA stability and translation ( through protein output ) . We determined that Dhh1-MS2 was functionally active , as it was able to complement a strain deleted for endogenous DHH1 ( i . e . dhh1Δ ) in assays for mRNA decapping ( unpublished data ) . We first evaluated whether Dhh1 altered mRNA decay when tethered to the 3′ UTR of a reporter mRNA . Wild-type ( WT ) cells expressing either MS2 or Dhh1-MS2 were evaluated for degradation of co-expressed MFA2 reporter mRNA . Importantly , reporter mRNAs are expressed from the regulatable GAL1 promoter , thereby permitting repression of reporter mRNA transcription and measurement of mRNA decay [13] . Cells were grown in the presence of galactose to induce reporter mRNA expression and , upon reaching mid-log phase , transcription was rapidly inhibited by replacing the media with glucose-containing media . Cells were harvested at indicated times and RNA isolated and analyzed by northern blot . As shown in Figure 1B , MFA2 reporter mRNA is destabilized by Dhh1 tethered to its 3′ UTR . Specifically , the half-life of MFA2 mRNA was reduced 2-fold by Dhh1-MS2 versus MS2 alone ( 3 . 4 min versus 6 . 3 min , respectively ) . Moreover , destabilization of the reporter mRNA required the MS2 binding sites , as MFA2 mRNA lacking the sites decayed with a half-life of approximately 6 min , similar to endogenously expressed MFA2 mRNA ( [13]; unpublished data ) . These results establish that Dhh1 , when associated with an mRNA through binding to its 3′ UTR , can accelerate the decay rate of the mRNA . MFA2 mRNA is inherently unstable and its degradation is particularly sensitive to alterations in mRNA decapping [12] . We therefore evaluated whether the destabilization of MFA2 reporter mRNA by tethered Dhh1 was mediated through changes in mRNA decapping rate . MFA2 reporter mRNA decay was measured in the presence of either MS2 or Dhh1-MS2 in cells lacking mRNA decapping activity ( i . e . dcp2Δ ) . MFA2 reporter mRNA in the presence of MS2 coat protein alone was dramatically stabilized by the absence of Dcp2 , similar to previous observations for endogenously expressed MFA2 mRNA ( Figure 1C; [2] ) . In contrast to our observation in wild-type cells , Dhh1-MS2 failed to lead to destabilization of MFA2 reporter mRNA in the absence of DCP2 , and the decay rate was essentially identical to that observed in cells expressing MS2 ( Figure 1C ) . These results indicate that Dhh1 destabilizes mRNA through a step at or before mRNA decapping when associated by tethering . We next set out to evaluate if Dhh1 can function as a translational repressor independent of its ability to promote mRNA decapping . To facilitate measurement of protein expression , MFA2 and PGK1 reporter mRNAs were generated in which their ORF was replaced with that of GFP ( Figure 1A; M/GFP and P/GFP , respectively ) . Wild-type cells harboring either MS2 coat protein alone or Dhh1-MS2 and either M/GFP or P/GFP reporter genes were evaluated for GFP protein expression by Western blot analysis . As shown in Figure 1D , Dhh1-MS2 caused a 50%–80% reduction in protein expression when tethered to reporter mRNAs as compared to MS2 alone . Considering the observation that tethered Dhh1 also promotes mRNA decay ( Figure 1B ) , one simple interpretation is that the reduced GFP protein level observed here is a consequence of reduced mRNA levels . To uncouple mRNA decay from a possible role for Dhh1 in repressing translation of the reporter mRNA , we repeated this analysis in the dcp2Δ strain , where tethering of Dhh1 did not alter mRNA decay rates ( Figure 1C ) . In these cells , Dhh1-MS2 still mediated a dramatic decrease in GFP protein expression from both reporters ( Figure 1D; GFP levels reduced 60%–70% ) . These data demonstrate that Dhh1 promotes repression of mRNA translation independent of promoting mRNA decapping when tethered to an mRNA , this is in agreement with recently published work [56] . Dhh1 has documented genetic and physical interactions with the deadenylase complex that , as the first step in mRNA degradation , removes the poly ( A ) tail from the mRNA [2] , [5] . To establish whether tethering of Dhh1 modulates translational repression by simply recruiting the deadenylase to the mRNA and thereby facilitating poly ( A ) tail removal , we evaluated the effect of Dhh1-MS2 on M/GFP reporter mRNA translation in cells lacking CCR4 ( i . e . ccr4Δ ) , the gene expressing the catalytic subunit of the deadenylase complex [2] . We observed that similar to wild-type cells , tethering of Dhh1 facilitated translational repression of M/GFP mRNA in cells lacking CCR4 ( Figure S5 ) , demonstrating that Dhh1 does not accelerate translational repression through removal of the poly ( A ) tail . Finally , we established whether the function of Dhh1 in our assays requires a functional DEAD-box protein domain . Dhh1-MS2 in which key functional residues of the DEAD-box motif were mutated ( DEAD to AAAD ) was unable to reduce M/GFP reporter mRNA levels or GFP protein expression ( Figure S1 ) , in contrast to our observations for Dhh1-MS2 ( Figures 1 and S1 ) . These results demonstrate that Dhh1-MS2 requires the DEAD-box for function , similar to observations for endogenously expressed Dhh1 [31] . Having established a robust assay to monitor the role of Dhh1 in repressing mRNA translation , we next set out to investigate the specific step of translation altered by Dhh1 function . Previous work from several labs suggested Dhh1 and its orthologs limit translation initiation prior to formation of the 48S pre-initiation complex [12] , possibly by antagonizing eIF4E binding to the mRNA 5′ cap [16] , [17] , [32] . If Dhh1 indeed controls translation by blocking eIF4E function or 48S complex formation , loss of eIF4E or eIF3 function would be predicted to abrogate observed effects of tethered Dhh1 on GFP expression . Temperature-sensitive alleles of CDC33 ( cdc33-1 , expressing eIF4E ) or PRT1 ( prt1-1 , expressing eIF3b ) inactivate protein function and reduce mRNA translation to less that 5% of that observed in wild-type cells at the restrictive growth temperature [33] , [34] . Importantly , residual mRNA translation allowed by these mutant alleles is required to be able to observe changes in mRNA translation of reporter mRNA . We were unable to use GFP protein levels to monitor changes in mRNA translation , however , since the 1-h incubation at the restrictive growth temperature sufficient to inactivate eIF4E or eIF3 function is short relative to the stability of GFP protein ( ∼7 h ) [35] . Therefore , mRNA levels were used to reflect the translation status of the mRNA . This method to evaluate mRNA translation has been used previously [14] and is consistent with our observation that the function of Dhh1 on mRNA is primarily at the level of translation , and that mRNA decay represents a secondary consequence of translational control ( Figure 1 ) . Isogenic wild-type or cdc33-1 cells co-expressing the M/GFP reporter with either MS2 alone or Dhh1-MS2 were grown to log phase at the permissive temperature ( 24°C ) and shifted to the restrictive temperature ( 37°C ) for 1 h prior to harvesting cells and isolating RNA for Northern blot analysis . Growth of the mutant strain at the restrictive temperature resulted in a 4-fold reduction in steady state levels of both M/GFP reporter mRNA and endogenous PGK1 mRNA in cells also expressing MS2 coat protein ( Figure 2A , compare lanes 1 and 3 ) . These data are consistent with previous observations [14] and demonstrate inactivation of eIF4E function under these growth conditions . Wild-type cells expressing Dhh1-MS2 displayed a 2-fold reduction in M/GFP mRNA levels compared to cells expressing MS2 alone ( Figure 2A , compare lanes 1 and 2 ) , consistent with the 2-fold reduction in decay rates by tethered Dhh1 ( Figure 1 ) . Relative to MS2 alone , Dhh1-MS2 resulted in an approximate 2-fold reduction in M/GFP reporter mRNA levels in cdc33-1 cells expressing temperature-inactivated eIF4E ( Figure 2A , compare lanes 3 and 4 ) . These observations reveal that Dhh1 functions to robustly modulate reporter mRNA levels ( through repressing mRNA translation ) even in the absence of fully functional eIF4E and when translation initiation is severely abrogated , suggesting that Dhh1 does not function through modulating eIF4E activity . The Xenopus homolog of Dhh1 , Xp54 , fails to repress translation of a reporter mRNA initiated from an internal ribosome entry site ( IRES ) [18] . Considering that IRES-mediated initiation does not require the eIF3 translation initiation complex , we hypothesized that it may be the target of Dhh1 function in repressing mRNA translation . To determine if eIF3 function is required for Dhh1-mediated effects on mRNA , we utilized cells harboring a temperature-sensitive allele of the gene expressing eIF3b ( i . e . prt1-1 ) . Importantly , this mutation in eIF3b leads to a significant disruption of the entire eIF3 complex and its function [36] . In prt1-1 cells at the non-permissive temperature , endogenous PGK1 mRNA levels are reduced approximately 4-fold ( Figure 2B , lanes 1 and 3 ) , demonstrating reduced eIF3b function as observed by others [14] . Interestingly , M/GFP reporter mRNA levels are insensitive to inactivation of eIF3b , suggesting that eIF3b is dispensable for the observed translation and mRNA turnover of this mRNA . Despite this , in eIF3b mutant cells Dhh1-MS2 was observed to still reduce M/GFP mRNA levels to approximately 20% relative to tethering MS2 alone ( Figure 2B ) . This level of mRNA reduction is similar to that observed for Dhh1-MS2 in wild-type cells , indicating that Dhh1 function is unlikely through limiting the function of the eIF3 complex in promoting translation initiation . Finally , we tested whether Dhh1 could modulate mRNA levels or translation of a reporter mRNA when translation of the mRNA is restricted in cis . mRNA translation was inhibited by the inclusion of a strong RNA secondary structure ( i . e . stemloop; SL ) in the 5′ UTR of a PGK1 reporter that has been demonstrated to limit 48S ribosome scanning ( Figure 2C; SL-PGK1 ) [12] , [13] . The 5′ SL leads to reduced protein production from the PGK1 reporter encoding a Pgk1-HA protein chimera ( Figure 2D , compare lanes 1 and 3 where cells express MS2 alone ) . Indeed , when normalized to a loading control ( i . e . ribosomal protein Rpl5 ) , translation of SL-PGK1 mRNA is less than 10% of the same reporter lacking the 5′ SL . In the presence of Dhh1-MS2 , protein expression from both PGK1 and SL-PGK1 reporters was dramatically reduced relative to MS2 alone ( Figure 2D; compare lanes 1 and 2 and lanes 3 and 4 ) . Moreover , Dhh1-MS2 also led to a substantial decrease in steady state mRNA levels for both reporters ( Figure 2E ) . These results demonstrate that despite an impairment in translation initiation at the level of ribosome scanning , Dhh1's function in inhibiting protein expression ( and subsequently mRNA abundance ) is not abrogated , and is as robust as that observed for reporter mRNAs undergoing translation in wild-type cells or in the absence of impediments presented by RNA structure . Together , these data indicate that repression of mRNA translation by Dhh1 is not mediated through modulation of eIF4E or eIF3 complex function , or 48S ribosome scanning . To further investigate the step of mRNA translation inhibited by Dhh1 , the association of reporter mRNAs with ribosomes was monitored . Sucrose density centrifugation represents a powerful and unbiased biochemical technique used for decades to inspect perturbations in the various steps of translation . We evaluated M/GFP reporter mRNA in cells co-expressing either MS2 alone or Dhh1-MS2 . Based on the loading of few ribosomes ( Figure 3A and 3B; MS2 ) , M/GFP reporter mRNA is ideally suited to observe changes in density based on alteration of its association with ribosomes . Mutant cells lacking mRNA decapping activity ( i . e . dcp2Δ ) were utilized to facilitate analysis of the effect of tethered Dhh1 on translation independent from secondary effects on mRNA turnover ( Figure 1 ) . Cell extracts were layered on sucrose gradients and polyribosome complexes were separated by velocity sedimentation . During fractionation , absorbance at 254 nm was measured and “polyribosome traces” were generated ( see Figure 3A ) . Total RNA was isolated from gradient fractions and M/GFP reporter mRNA was detected by northern blot . The polyribosome distribution of M/GFP mRNA from dcp2Δ cells expressing MS2 alone indicated that the mRNA associates predominantly with between 1 and 5 ribosomes ( Figure 3A ) . In dramatic contrast , in the presence of Dhh1-MS2 , the sedimentation of M/GFP mRNA shifted to a region deep within the gradient , consistent with heavy polyribosomes ( Figure 3A ) . Importantly , Dhh1-MS2 did not lead to the accumulation of ribosome-free M/GFP mRNA detectable by sedimentation in non-ribosomal fractions 1 or 2 , as would have been expected if tethered Dhh1 was inhibiting translation at initiation . The detection of M/GFP mRNA in dense regions of the gradient when Dhh1 is tethered is consistent with but not conclusive evidence that ribosomes are abundantly associated with the mRNA . To directly determine the association of M/GFP mRNA with ribosomes , ribosomes were affinity purified from cell extracts and the associated RNA measured by qRT-PCR [37] . Yeast cells expressing a C-terminally tagged version of ribosomal protein Rpl16a ( Rpl16a-ZZ ) [37] were mutated to delete DCP2 and then were used in subsequent experiments . Extracts were prepared from these cells expressing M/GFP reporter mRNA and either MS2 or Dhh1-MS2 and ribosomes immunoprecipitated using an anti-TAP antibody ( see Materials and Methods ) . Visualization of co-purified RNA separated by agarose gel electrophoresis confirmed recovery of 18S rRNA from lysates containing tagged Rpl16a compared to an untagged control ( Figure 3C ) . The association of specific mRNAs within the co-purified material was measured by qRT-PCR and normalized to the level of U1 snRNA , a non-translated RNA that associates relatively inefficiently with ribosomes [37] . We observed that both endogenous and reporter mRNA can be efficiently co-purified relative to U1 snRNA using this approach ( Figure 3D ) . Moreover , reporter mRNA from cells expressing Dhh1-MS2 is co-purified to a similar extent as MS2 alone ( Figure 3D; M/GFP mRNA; compare red and black bars ) . Importantly , co-purification of these mRNA targets is several hundred-fold enriched over that detected from similar experiments using lysates with untagged Rpl16a , indicating the specificity of the method ( unpublished data ) . Our data suggest two important things . First , tethered Dhh1 does not lead to a large-scale dissociation of ribosomes from the mRNA , and second , the sedimentation of M/GFP reporter mRNA deep in polyribosome gradients ( Figure 3A ) must be due , in part , to its association with ribosomes . This latter observation is also inconsistent with the sedimentation of a large mRNP aggregate that lacks an association with ribosomes , such as P bodies [20] . To more rigorously establish that the dense sedimentation of M/GFP mRNA in the presence of Dhh1-MS2 represents ribosome-associated material , ribosomes were affinity purified from cell lysates as described above , and the co-purified material then subjected to sucrose density gradient sedimentation . Gradient fractions were collected and the abundance of reporter mRNA throughout the fractions measured by qRT-PCR . The ratio of mRNA present in gradient fractions from cells expressing Dhh1-MS2 versus MS2 was determined . M/GFP reporter mRNA showed a significant overrepresentation in dense polyribosome fractions in the presence of Dhh1-MS2 ( Figure 3F; fractions 13–16 ) and a coordinate underrepresentation in the remainder of the fractions ( Figure 3F; fractions 1–12 ) . This observation is in strong correlation to that observed for this reporter mRNA subject directly to gradient sedimentation and analyzed by Northern blot ( Figure 3A and quantified in 3E ) . The slightly reduced enrichment of reporter mRNA in dense gradient fractions in the presence of Dhh1-MS2 from cell lysates that were affinity purified reflects more efficient recovery of light polyribosomes over heavy polysomes by this approach ( Figure S6; [37] ) . Notwithstanding , tethered Dhh1 causes the increased sedimentation of reporter mRNA in sucrose gradients and this material is clearly associated with ribosomes . Moreover , mRNA repressed in their translation by tethered Dhh1 appear to be associated with a larger number of ribosomes than during their basal metabolism and may indicate that Dhh1 functions to limit translation at some late step , perhaps at elongation , termination , or the poorly characterized ribosome recycling step . Our data utilizing tethered-function analysis to analyze Dhh1 suggests that the tethered protein represses mRNA translation at a step after initiation and that it inhibits disassociation of ribosomes from mRNA . We predicted that if endogenously expressed Dhh1 were performing the same function , Dhh1 should be found associated with polyribosomes . We and others have documented , however , that Dhh1 sediments with the soluble RNP in sucrose gradients [12] , [38] . We reasoned that the association of Dhh1 with polyribosomes in cells with active decay machinery and minimal cues for translational repression ( i . e . mid-log phase cells undergoing exponential growth ) may be transient and difficult to detect biochemically . To evaluate this hypothesis , cells were treated with formaldehyde in vivo to promote crosslinking and stabilize Dhh1-polysome complexes [39] . The sedimentation of Dhh1 with polysomes and other translation-associated mRNPs was then evaluated by sucrose gradient sedimentation . For this analysis , dhh1Δ cells expressing a plasmid-encoded , epitope-tagged Dhh1 protein ( HBHT-Dhh1 , [40] ) were utilized . Importantly , HBHT-Dhh1 is fully functional and complements dhh1Δ cells for growth and the metabolism of EDC1 mRNA ( Figure S2 ) . As shown in Figure 4A , in the absence of formaldehyde , HBHT-Dhh1 fails to co-sediment with polyribosomes , as previously observed [12] , [38] . In contrast , after mild crosslinking , HBHT-Dhh1 is present in heavy sucrose gradient fractions , suggesting that it co-sediments with polyribosomes . Treatment of cell extracts with RNase A prior to centrifugation abrogates the co-sedimentation pattern , indicating that the association of Dhh1 with dense material on sucrose gradients is mediated by RNA contacts , consistent with its association with polyribosomes and its ability to bind RNA [41] . Our evidence indicates that tethered Dhh1 limits translation at a step after initiation and increases the sedimentation of reporter mRNA in sucrose gradients ( Figure 3 ) and wild-type Dhh1 is associated with polyribosomes ( Figure 4A ) . Based on these observations , we hypothesized that wild-type Dhh1 may also play a role in inhibiting ribosome elongation , termination , and/or ribosome recycling . In any case , it would be predicted that after a block in translation initiation , Dhh1-bound mRNA would retain a prolonged association with ribosomes . To measure the association of Dhh1 with polyribosomes after inhibition of translation , cells were treated with 1 M sodium chloride for 10 min prior to harvesting and polysome analysis . Exposure of cells to high salinity inhibits translation and results in ribosome run-off from mRNAs and loss of polyribosomes as measured by sucrose gradient centrifugation [42] . Even in the presence of low levels of formaldehyde , treatment of cells expressing HBHT-tagged Dhh1 led to a significant loss of polysomes , as anticipated ( Figure 4B ) [42] . Polysome analysis followed by Western blot demonstrated that HBHT-Dhh1 remained predominantly associated with dense sucrose gradient fractions after inhibition of translation by high salt ( Figure 4C ) . In contrast , Dhh1 harboring a mutation in the DEAD-box that abrogates Dhh1 function in repressing translation ( Figure S1 ) fails to remain associated with polyribosomes under salt stress ( Figure 4D ) . Taken together , these data support that Dhh1 associates with polyribosomes and that it acts to restrict the dissociation of ribosomes from polyribosomes as measured by in vivo ribosome run-off analysis . To confirm that the association of HBHT-Dhh1 with dense sucrose gradient fractions represents its association with polyribosomes , ribosomes were affinity purified from cells grown in the presence or absence of salt stress . Consistent with the co-sedimentation of Dhh1 with polyribosomes ( Figure 4A ) , HBHT-Dhh1 co-purifies with ribosomes ( Figure 4E ) . Moreover , after inhibition of translation with high salt , Dhh1 maintains an association with ribosomes ( Figure 4F ) , consistent with its co-sedimentation with polysomes by sucrose gradient centrifugation . The observation that Dhh1 functions to limit ribosome run-off is consistent with Dhh1 inhibiting a step in translation subsequent to initiation and perhaps through limiting translation elongation . Moreover , as a consequence of Dhh1 function , mRNA decapping rate is enhanced leading to accelerated turnover of the mRNA ( Figure 1 ) . We hypothesized that inhibition of translation elongation by other means might also lead to a stimulation of mRNA decapping rate . To test this idea , a stretch of rare codons that restrict ribosome elongation [25] was inserted 77% into the coding region of a PGK1 reporter gene ( PGK1RC77%; Figure 5A ) . The rare codons greatly reduced Pgk1 protein expression to roughly 10% of wild-type PGK1 reporter mRNA ( Figure S3 ) , demonstrating the inhibition of translation elongation . Importantly , PGK1 reporter mRNA harboring the rare codons remains a substrate for mRNA decapping and 5′–3′ mRNA decay and is not targeted for No-go decay , as deletion of DOM34 failed to significantly stabilize this reporter while deletion of factors important for 5′–3′ mRNA degradation significantly stabilized the mRNA [25] . Transcriptional shut-off analysis of both PGK1 and PGK1RC77% in wild-type cells shows a significant destabilization of the mRNA dependent upon the rare codon stretch ( Figure 5B ) . Specifically , the decay rate of PGK1RC77% mRNA is accelerated 3-fold versus PGK1 mRNA lacking the rare codons ( half-life of 9 min versus 27 min , respectively ) . These data demonstrate the inhibition of translation elongation can indeed elicit the acceleration of mRNA decapping . If Dhh1 functions exclusively to inhibit translation elongation , the limitation of translation elongation mediated by the rare codons should bypass the need for Dhh1 in its rapid turnover of the reporter mRNA . We repeated the decay analysis for both PGK1 and PGK1RC77% in cells in which DHH1 was deleted ( i . e . dhh1Δ ) . The decay of PGK1 reporter mRNA was unaffected in dhh1Δ cells ( Figure 5B and 5C ) , indicating that this mRNA is degraded in a Dhh1-independent manner . In contrast , the Lsm1–7 complex has a profound effect on PGK1 mRNA stability ( unpublished data ) . It is unclear why PGK1 reporter mRNA is not a substrate for Dhh1 activity , but it will be an important mRNA in further elucidating Dhh1 function . Notwithstanding , PGK1RC77% mRNA was stabilized 3-fold in dhh1Δ cells compared to WT ( Figure 5C ) , indicating that limiting ribosome movement on a reporter mRNA is not sufficient to bypass the requirement for Dhh1 function . Interestingly , the inhibition of ribosome elongation in cis does , instead , serve to render an otherwise Dhh1-insensitive mRNA into one that now responds to Dhh1 in the cell .
All mRNA succumbs to degradation; therefore , decay represents a default state in mRNA metabolism . The spectrum of mRNA half-lives observed for different mRNAs and in different cell types represents the acceleration or inhibition of the default rate of decay . One major factor that significantly contributes to the overall stability of an mRNA is its translatability [1] , [43] . Indeed , an inverse correlation has been established wherein efficiently translated mRNAs display longer half-lives while poorly translated mRNAs are generally unstable . Competition for binding at or near the mRNA 5′ 7-methyl cap between the translation initiation factor eIF4E and the catalytic peptide of the decapping complex , Dcp2 , is consistent with the observed inverse correlation between translation and mRNA decay . It is therefore generally assumed that modulating translational initiation is a key event in regulating the rate of mRNA decapping [14] , [15] . The DEAD-box RNA helicase Dhh1 and its homologues have been implicated as active stimulators of mRNA decapping through dissociation of the translation initiation complex from mRNA . Specifically , Dhh1 proteins have been proposed to block initiation by interfering with eIF4E function [16] , [32] or with eIF3-mediated 48S ribosomal complex assembly [12] , [18] . Our previous work appeared to support these ideas [12] . Deletion of DHH1 in combination with a second activator of mRNA decapping , PAT1 , prevented broad repression of mRNA translation in response to glucose deprivation as analyzed by polysome analysis [12] . At that time , glucose deprivation was believed to cause widespread inhibition of translation initiation [44] , and thus , our findings indicated that Dhh1 was required , in part , to modulate this process . Moreover , Dhh1 over-expression mediated a loss of bulk polysomes consistent with a general block to translation initiation . Finally , in vitro analysis of translation initiation complex assembly indicated that Dhh1 inhibited 48S complex formation on mRNA [12] . Advances in our understanding of mRNA metabolism call for new interpretations to previous observations . Recently , Arribere et al . showed that glucose deprivation leads to rapid and widespread degradation of most cellular mRNAs , rather than a general decrease in translation initiation [45] . The overall collapse in polyribosomes seen upon glucose deprivation is most likely a manifestation of this generalized decay phenomena . In our work from 2005 [12] , the RPL41a mRNA was used to illustrate that mRNAs relocated from polyribosomes to non-polyribosome fractions upon glucose deprivation and that decay was not affected . Indeed , as a ribosomal protein gene , RPL41a belongs to the small class of mRNA not degraded following cell stress [45] but does dissociate from polyribosomes upon stress . Further work from our lab revealed that mRNAs targeted for decapping are not devoid of ribosomes , but rather , decapping occurs co-translationally while the mRNA is still associated with ribosomes [24] , [25] . These observations highlight that a fundamental change in the association of an mRNA with ribosomes does not occur before mRNA decapping as previously hypothesized , but rather that mRNA decapping is co-translational . Our findings presented here demonstrate that Dhh1 functions to repress mRNA translation , independent of any additional effect on promoting mRNA decapping ( Figure 1 ) . Moreover , Dhh1 functions at a step late in mRNA translation . Our data indicate that Dhh1 does not act through inhibiting eIF4E or eIF3 function ( Figure 2A and 2B ) . Dramatically , when Dhh1 is tethered to a reporter , mRNA translation is repressed yet the mRNA co-sediments with denser polysomes that represent an increased association of the mRNA with ribosomes ( Figure 3 ) . Consistent with this , endogenous Dhh1 associates with polysomes , albeit in a transient manner . Finally , we show using saline-induced inhibition of translation initiation that Dhh1-polyribosome complexes dissociate from mRNA ( i . e . run off ) slowly ( Figure 4 ) . Together , these data demonstrate that Dhh1 is a bona fide translational repressor in vivo and that its function is consistent with a role in slowing ribosome movement on mRNA . The function of Dhh1 in regulating translation post-initiation is consistent with phenomena observed in several additional biological contexts . First , two developmentally regulated mRNAs repressed on polyribosomes in Drosophila embryos , oskar and nanos , are inhibited for translation at some level by the Dhh1-homolog Me31b [7] , [46]–[48] . Human KRAS mRNA is repressed on polyribosomes by let-7 miRNA in human cells [49] , and this repression is partially attributed to RCK/p54 [10] . Interestingly , ribosome run-off of let-7-targeted KRAS mRNA occurs more slowly in response to a stress-induced translation initiation block [49] , consistent with the repressed KRAS mRNP also being associated with slowly moving ribosomes . Finally , the documented purification of ribosomes with Dhh1 as well as its co-purification of translation elongation factor 1a in an RNA-independent manner [38] support Dhh1 as a repressor of a late step in mRNA translation . Interestingly , Fragile X Mental Retardation Protein ( FMRP ) , a polysome-associated neuronal RNA binding protein with interactions with Me31b [9] , was also recently found to regulate translation by inducing stalling of ribosomes on target mRNAs [50] . One potentially unifying theory of the data we have presented previously [12] and our current findings is that Dhh1 directly affects the function of the 40S ribosomal subunit . Indeed , we and others have observed that Dhh1 binds ribosomes [38] . Moreover , Dhh1 represses translation in vitro of an mRNA harboring the Cricket Paralysis Virus IRES , which requires only 40S ribosomes to initiate translation [12] . The context upon which Dhh1 binds to the 40S ribosomal subunit might affect which step in translation that appears to be inhibited ( Figure 6 ) . Interaction between Dhh1 and free 40S subunits could influence translation at early steps and manifest as an initiation block . This mechanism might be occurring both in vitro and during Dhh1 over-expression in cells [12] . In the context of an actively translating mRNA , however , Dhh1 interaction with 40S subunits might impede ribosome movement on mRNA as we have observed and implies a role for Dhh1 in inhibiting translation either during elongation , termination , or ribosome recycling . Additional experiments will be needed to define precisely how Dhh1 functions mechanistically , but the two sets of data need not be mutually exclusive . Our data here suggest that Dhh1 may also function as a sensor for slowed translation elongation . Reducing ribosome elongation rate by the insertion of rare codons in a coding region of a reporter mRNA renders the mRNA unstable ( Figure 5 ) and converts the mRNA into a substrate for Dhh1-mediated mRNA decay . This observation indicates that the accelerated decay in response to slowed ribosome elongation requires Dhh1 . Interestingly , dhh1Δ cells also demonstrate an increased sensitivity to three general inhibitors of translation elongation ( Figure S4 ) , suggesting that in the absence of Dhh1 , cells have a reduced ability to resolve the effects of a general inhibition of ribosome movement . The transient interaction of Dhh1 with polyribosomes ( Figure 4 ) may reflect rapid sampling of polyribosome complexes by Dhh1 , a common theme for biological sensors . The role of Dhh1 as both a sensor of slowed ribosome movement and a mediator of translational repression is reminiscent of the function of another ATP-dependent RNA helicase , Upf1 , in the decay of nonsense-containing mRNA . Upf1 is required for the recognition of aberrant translation termination events and in response to this event , mediates both translational repression and accelerated decapping of the mRNA [51] . For Dhh1-like proteins , one key regulatory event that may induce activity is removal of the mRNA 3′ poly ( A ) tail [17] , [32] . Deadenlyation leads to the loss of poly ( A ) binding protein ( Pab1 ) association with the mRNA and dramatic changes in the translational status of the mRNA are predicted to occur at many different levels , including elongation , termination and ribosome recycling [52]–[55] . In this light , we postulate that mRNA decapping serves an important role , preventing further translation from translationally impaired transcripts .
Yeast strains are listed in Table S1 . Unless otherwise noted , all strains were grown at 24°C in synthetic media with the appropriate amino acids and either 2% galactose/1% sucrose , 4% glucose ( for shutting off the GAL1 UAS ) , or 2% glucose as appropriate . All cells were harvested at mid-log phase ( OD600 = 0 . 4–0 . 55 ) . Temperature-sensitive translation initiation mutant cells ( yJC102 , 104 , 1011 , or 1012 ) were shifted to the non-permissive temperature ( 37°C ) for 1 h before harvesting . Cell stress experiments in Figure 4 were carried out by growing cells to mid-log phase , centrifuging the cells , and resuspending the cells in media with or without 1 M NaCl , then immediately adding formaldehyde as described below . Details in Text S1 and Table S2 . Cells ( yJC151 , 327 , or 330 ) expressing the appropriate plasmids were grown to mid-log phase in synthetic media containing 2% galactose/1% sucrose to allow expression of reporter mRNAs , then were centrifuged and resuspended in synthetic media without sugar . The 0 min time point was harvested , then glucose was added to a final concentration of 4% to shut off transcription . Cells were harvested at the time points indicated in each figure , then RNA was isolated by glass bead lysis followed by phenol/chloroform extraction and ethanol precipitation . 20–40 µg of total RNA from each time point were separated on 1 . 4% agarose-formaldehyde gels , transferred to nylon membranes , and probed overnight with 32P end-labeled oligonucleotides ( listed in Table S2 ) . RNAs were probed for using an oligonucleotide antisense to the MS2 binding sites ( oJC1006 ) , PGK1 ( oJC357 ) , EDC1 ( oJC221 ) , or SCR1 ( oRP100 ) . Blots were exposed to PhosphorImager screens , scanned using a Storm 820 scanner , and quantified with ImageQuant software . Cells were grown to mid-log phase and harvested . Protein was isolated by resuspending cells in 200 µL 5 M urea , heating to 95°C for 2 min , vortexing cells with glass beads for lysis , adding 500 µL solution A ( 125 mM Tris-HCl pH 6 . 8 , 2% SDS ) , vortexing 1 min , heating to 95°C for 2 min , and finally clearing extracts by centrifugation at 13 , 300 rpm for 2 min . Equivalent OD280 of extract was loaded onto 10% SDS polyacrylamide gels . Protein was transferred to PVDF membrane and blotted for various proteins ( anti-HA , Covance; anti-Pab1 , EnCor Biotechnology; anti-Rpl5; anti-Pgk1 , Invitrogen; anti-RGS-His , Qiagen ) . Detection was carried out using Amersham ECL kit and exposing blots to Blue Ultra AutoRad film ( ISC Bioexpress ) . Quantification was carried out by scanning the film and using ImageJ software . Cells were harvested in 100 µg/mL cycloheximide . Cells used in Figure 4 were crosslinked at a final concentration of 0 . 25% formaldehyde for 5 min , then treated with 125 mM glycine for 5 min ( Figures 4C through 4F ) or 10 min ( Figure 4A ) to quench crosslinking . Cells were then lysed into 1× lysis buffer ( 10 mM Tris pH 7 . 4 , 100 mM NaCl , 30 mM MgCl2 , 0 . 5 mg/mL heparin , 1 mM DTT , 100 µg/mL cycloheximide ) by vortexing with glass beads , and cleared using the hot needle puncture method followed by centrifugation at 2 , 000 rpm for 2 min at 4°C , then incubated in 1% Triton X-100 for 5 min on ice . In Figure 3A , 20 OD260 units were loaded on 15%–45% ( w/w ) sucrose gradients prepared on a Biocomp Gradient Master in 1× gradient buffer ( 50 mM Tris-acetate pH 7 . 0 , 50 mM NH4Cl , 12 mM MgCl2 , 1 mM DTT ) and centrifuged at 41 , 000 rpm for 1 h and 13 min at 4°C in a Sw41Ti rotor . Gradients were fractionated using a Brandel Fractionation System and an Isco UA-6 ultraviolet detector . Fractions were precipitated overnight at −20°C using 2 volumes 95% ethanol . RNA/protein was pelleted at 14 , 000 rpm for 30 min , then pellets were resuspended in 500 µL LET ( 25 mM Tris pH 8 . 0 , 100 mM LiCl , 20 mM EDTA ) with 1% SDS . Fractions were then extracted once with phenol/LET , once with phenol/chloroform/LET , and then were precipitated with one-tenth volume of 7 . 5 M CH3COONH4 and 2 volumes 95% ethanol . RNA pellets were recovered by centrifugation at 14 , 000 rpm for 30 min . Pellets were washed once with 700 µL 75% ethanol , air dried , and resuspended in 1× sample buffer ( 200 mM MOPS pH = 7 . 0 , 50 mM sodium acetate , 12 . 5 mM EDTA , 3 . 33% formaldehyde , 0 . 4 mg/mL ethidium bromide ) , and then samples were heated to 65°C for 10 min to denature RNA . The entire sample was then loaded on 1 . 4% agarose-formaldehyde gels and Northern analysis carried out as above . For Western blot analysis of protein from sucrose gradients , fractions were precipitated with a final concentration of 10% TCA , pellets were washed with 80% acetone , then allowed to air dry . Pellets were resuspended in 1× SDS-PAGE loading buffer , boiled , and loaded on 10% SDS polyacrylamide gels , then processed as in the section on Western blots . dcp2Δ cells expressing a chromosomally ZZ-tagged version of Rpl16a ( yJC1141 ) were grown to mid-log phase and harvested . Procedures were adapted from [37] . Cell lysis was performed as for polyribosome analysis by vortexing with glass beads in 1× lysis buffer without heparin . Samples were brought to 300 µL with 1× lysis buffer . Samples were then brought up to 592 µL with 2× binding buffer ( 100 mM Tris-HCl pH = 7 . 5 , 24 mM Mg ( CH3COO ) 2 , 1 mM DTT , 100 µg/mL cycloheximide ) . Lysates were incubated at 4°C overnight with 4 µg anti-TAP antibody ( Open Biosystems ) . The next morning , 1 . 5 mg protein-G Dynabeads ( Invitrogen ) were washed 3 times in a mixture of equal parts 1× lysis buffer and 2× binding buffer . The lysate from the night before was then incubated with protein-G Dynabeads for 1 h at 4°C . Pellets were washed 4 times in IXA-500 buffer ( 50 mM Tris-HCl pH = 7 . 5 , 500 mM KCl , 12 mM Mg ( CH3COO ) 2 , 1 mM DTT , 100 µg/mL cycloheximide ) and RNA/protein was eluted with elution buffer ( 50 mM Tris-HCl pH = 7 . 5 , 0 . 5% SDS , 50 mM EDTA ( pH = 8 . 0 ) ) at 95°C for 5 min or TEV protease cleavage ( 100 U for 2 h in buffer C [20 mM Tris pH = 8 . 0 , 140 mM KCl , 2 mM MgCl2 , 5% glycerol , 0 . 5 mM DTT , 100 µg/mL cycloheximide] ) for loading onto gradients ( Figure 3F ) . RNA was isolated from one-tenth of the input or from the entire pelleted material by two phenol/chloroform extractions followed by chloroform extraction , then precipitated by sodium chloride and isopropanol . RNA was treated with 40 units of Roche DNase I , then extracted once with phenol/chloroform/LET and precipitated with sodium chloride and isopropanol , then resuspended in 15 µL of DEPC-treated dH2O . Reverse transcription was carried out using First Strand cDNA Synthesis Kit for Real-Time PCR from USB using random primers ( or oJC1470 for 25S rRNA ) and 1 µL of either a 32-fold dilution of input RNA from above or a 4-fold dilution of eluted RNA from each immunoprecipitation . qPCR was carried out using VeriQuest SYBR Green Master Mix ( USB ) in a StepOne Real Time PCR system ( Applied Biosystems ) and the following oligonucleotides: GFP , oJC1240 , 1241; MFA2 , oJC983 , 984; PGK1 , oJC985 , 986; U1 , oJC989 , 990; 25S rRNA , oJC1470 , 1471 . Relative differences between samples were calculated using the ΔΔCt method . A dilution series for each target ensured that we were within the linear range of the assay ( unpublished data ) . | Translation of mRNA into protein and turnover of mRNA are two points at which cells can exert regulatory control of gene expression , thereby ensuring that the protein products are present in cells and tissues at the appropriate time and place . The DDX6 family of DEAD box helicases , exemplified by the yeast protein Dhh1 , is a group of well-conserved eukaryotic proteins that regulate translation and mRNA decay . As DDX6 proteins are known to be important for diverse processes such as cellular stress responses , early embryonic development , and replication of some viruses , understanding their mechanism of action could be of broad significance to many fields . Previous studies suggest that Dhh1 and other DDX6-family proteins mainly regulate translation at the initiation stage , triggering sequestration and/or decapping of the mRNA . Our work expands the potential functions of Dhh1 , showing that Dhh1 is also capable of inhibiting translation at later stages when ribosomes are already loaded onto mRNAs . This extended function for Dhh1 allows a more robust translational control , as inhibition at a late stage of translation can provide immediate stoppage of protein production , as well as affording the potential for storing mRNA already primed and loaded with ribosomes for subsequent rapid re-utilization . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"biochemistry",
"genetics",
"biology",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
]
| 2012 | The DEAD-Box Protein Dhh1 Promotes Decapping by Slowing Ribosome Movement |
The rise of multi-drug resistant ( MDR ) and extensively drug resistant ( XDR ) tuberculosis around the world , including in industrialized nations , poses a great threat to human health and defines a need to develop new , effective and inexpensive anti-tubercular agents . Previously we developed a chemical systems biology approach to identify off-targets of major pharmaceuticals on a proteome-wide scale . In this paper we further demonstrate the value of this approach through the discovery that existing commercially available drugs , prescribed for the treatment of Parkinson's disease , have the potential to treat MDR and XDR tuberculosis . These drugs , entacapone and tolcapone , are predicted to bind to the enzyme InhA and directly inhibit substrate binding . The prediction is validated by in vitro and InhA kinetic assays using tablets of Comtan , whose active component is entacapone . The minimal inhibition concentration ( MIC99 ) of entacapone for Mycobacterium tuberculosis ( M . tuberculosis ) is approximately 260 . 0 µM , well below the toxicity concentration determined by an in vitro cytotoxicity model using a human neuroblastoma cell line . Moreover , kinetic assays indicate that Comtan inhibits InhA activity by 47 . 0% at an entacapone concentration of approximately 80 µM . Thus the active component in Comtan represents a promising lead compound for developing a new class of anti-tubercular therapeutics with excellent safety profiles . More generally , the protocol described in this paper can be included in a drug discovery pipeline in an effort to discover novel drug leads with desired safety profiles , and therefore accelerate the development of new drugs .
Tuberculosis , which is caused by the bacterial pathogen Mycobacterium tuberculosis ( M . tuberculosis ) , is a leading cause of mortality among infectious diseases . It has been estimated by the World Health Organization ( WHO ) that almost one-third of the world's population , around 2 billion people , is infected with the disease [1] . Every year , more than 8 million people develop an active form of the disease , which subsequently claims the lives of nearly 2 million . This translates to over 4 , 900 deaths per day , and more than 95% of these are in developing countries [2] . In 2002 , the WHO estimated that if the worldwide spread of tuberculosis was left unchecked , then the disease would be responsible for approximately 36 million more deaths by the year 2020 . Despite the current global situation , anti-tubercular drugs have remained largely unchanged over the last four decades [3] . The widespread use of these agents , and the time needed to remove infection , has promoted the emergence of resistant M . tuberculosis strains . Multi-drug resistant tuberculosis ( MDR-TB ) is defined as resistance to the first-line drugs isoniazid and rifampin . The effective treatment of MDR-TB necessitates the long-term use of second-line drug combinations , an unfortunate consequence of which is the emergence of extensively drug resistant tuberculosis ( XDR-TB ) – M . tuberculosis strains that are resistant to isoniazid plus rifampin , as well as key second-line drugs , such as ciprofloxacin and moxifloxacin . XDR-TB is extremely difficult to treat because the only remaining drug classes exhibit very low potency and high toxicity . The rise of XDR-TB around the world , including in industrialized nations , imposes a great threat on human health , therefore emphasizing the need to identify new anti-tubercular agents as an urgent priority [4] . Currently , anti-infective therapeutics are discovered and developed by either de novo strategies , or through the extension of available chemical compounds that target protein families with the same or similar structures and functions . De novo drug discovery involves the use of high throughput screening techniques to identify new compounds , both synthetic and natural , as novel drugs . Unfortunately , this approach has yielded very few successes in the field of anti-infective drug discovery [5] . Indeed , the progression from early-stage biochemical hits to robust lead compounds is commonly an unfruitful process . The identification of both molecular targets that are essential for the survival of the pathogen , and compounds that are active on intact cells , is a challenging task . Even more formidable , however , is the requirement for appropriate potency levels and suitable pharmacokinetics , in order to achieve efficacy in small animal disease models [5] . These challenges are reflected in the high costs involved in bringing new drugs to market . In fact , it has been estimated that the successful launch of a single new drug costs more than US$800 million [6] . Two alternative drug discovery strategies that circumvent some of the challenges associated with de novo drug discovery are the label extension and ‘piggy-back’ strategies , both of which are widely employed for the discovery of novel therapeutics to treat tropical diseases . Label extension is a fast-track approach that involves the extension of the indications of an existing treatment to another disease . Some of the most important anti-parasitic drugs in use today , such as praziquantel for schistosomiasis , were derived from the label extension process . The major advantages of label extension are the significant reductions in cost and time to market that can be achieved . Alternatively , when a molecular target that is present in a pathogen is under investigation for other commercial indications , it is possible to adopt the ‘piggy-back’ strategy by utilizing the identified chemical starting points . Examples of this approach include the anti-malarial screening of a lead series of cysteine protease inhibitors for the treatment of osteoporosis , and histone deacetylase inhibitors for use in cancer chemotherapy [5] . One of the main aims of drug discovery is to develop safe and effective therapeutic agents through the optimization of binding to a specific protein target . In this way , undesirable effects resulting from side scatter pharmacology are minimized . However , the recent and rapid completion of numerous genome sequencing projects has revealed that proteins involved in entirely different biochemical pathways , and even residing in different tissues and organs , may possess functional binding pockets with similar shapes and physiochemical properties [7] . Therefore , chemical matter for one target could be considered as the basis for leads for an entirely different target . Recent work on large scale mapping of polypharmacology interactions by Paolini et al . [8] revealed the extent of promiscuity of drugs and leads across the proteome . They discovered that around 35% of 276 , 122 active compounds in their database had observed activity for more than one target . Whilst the majority of these promiscuous compounds were found to be active against targets within the same gene family , a significant number ( around 25% ) had recorded activity across different gene families . The finding that so many drugs interact with more than one target provided the rationale behind the selective optimization of side activities ( SOSA ) approach recently developed by Wermuth [9] , [10] . The SOSA approach involves the use of old drugs for new pharmacological targets , which is a valuable concept considering the finite number of small molecules that can be safely administered to humans . The process itself involves screening a limited number of structurally diverse drug molecules , and then optimizing the hits so that they show a stronger affinity for the new target and a weaker affinity for the original target ( s ) . In this way , it is possible to derive a whole panel of new active molecules from a single marketed drug . Since the screened drug molecules already have known safety and bioavailability in humans , the overall time and cost of drug discovery is significantly reduced when compared with de novo strategies . We have developed a novel computational strategy to identify off-targets of major pharmaceuticals on a proteome-wide scale [11]–[15] . Our methodology extends the scope of the SOSA concept effectively and systematically across gene families , and is more likely to be successful in achieving the ultimate goal of providing new drugs from old ones . Our chemical systems biology approach proceeds as follows: Our approach essentially explores complex protein-ligand interaction networks on a proteome-wide scale . The lead compound can be discovered from all drug targets across different gene families . Moreover , lead optimization can focus on compounds with excellent safety profiles and known clinical outcomes . In this way , our approach has the potential to increase the rate of successful drug discovery and development , whilst reducing the costs involved . In the present study we demonstrate the efficiency and efficacy of our chemical systems biology approach through the discovery of safe chemical compounds with the potential to treat MDR-TB and XDR-TB . The identified compounds are entacapone and tolcapone . These drugs primarily target human catechol-O-methyltransferase ( COMT ) , which is involved in the breakdown of catecholamine neurotransmitters such as dopamine . They are used as adjuncts to treat Parkinson's disease by increasing the bioavailability of the primary drug levodopa , which is a substrate of COMT . Entacapone and tolcapone are predicted to inhibit M . tuberculosis enoyl-acyl carrier protein reductase ( InhA ) , which is essential for type II fatty acid biosynthesis and the subsequent synthesis of the bacterial cell wall [16] . InhA is the target of the anti-tubercular drugs isoniazid [17] and ethionamide [18] . Similar to newly developed direct InhA inhibitors [3] , [19]–[21] , entacapone and tolcapone require no enzymatic activation to bind InhA . Thus they may avoid the commonly observed resistance mechanism to isoniazid and ethionamide that is exhibited by many MDR strains . Our computational predictions have been partially validated by demonstrating that the entacapone drug tablet Comtan inhibits the growth of M . tuberculosis at the minimal inhibition concentration ( MIC99 ) of entacapone of 260 . 0 µM , well below the concentration leading to neuroblastoma cellular toxicity . The direct inhibition of InhA by entacapone is further confirmed by experimental enzyme kinetic assays , in which Comtan is shown to reduce InhA activity by up to 47% at the effective entacapone concentration of 80 µM . Since entacapone has an excellent safety profile with few side effects , it shows potential as a drug lead in the development of a new class of anti-tubercular therapeutics with favorable ADME/Tox properties .
Recently , Xie and Bourne developed a sequence order independent profile-profile alignment ( SOIPPA ) algorithm [11] , which they subsequently used to detect common binding sites among proteins unrelated in sequence and/or function . Their studies implied an evolutionary relationship between the NAD-binding Rossmann fold and several other fold classes , including the SAM-binding domain of the SAM-dependent methyltransferases , through similarities between their co-factor binding sites . It is interesting to note that both nicotinamide adenine dinucleotide ( NAD ) and S-adenosyl methionine ( SAM ) include adenine as a common molecular fragment . In fact , previous studies have shown that adenine binding pockets from proteins lacking significant homology will share common physiochemical properties [11] . These findings , plus the versatility of our method , form the basis for the present study . Entacapone and tolcapone are drugs that block the ligand binding site of COMT , a member of the SAM-dependent methyltransferase superfamily , in the presence of the SAM co-factor . They are used as adjuncts in Parkinson's disease therapy , to prevent the metabolism of levodopa to 3-O-methyldopa , therefore improving levodopa bioavailability and increasing its delivery to the brain [22] . The dopamine precursor levodopa has been the key drug for symptomatic treatment of Parkinson's disease for more than 30 years . Since COMT is a SAM-dependent methyltransferase , it is possible that it may possess a ligand binding pocket similar to those found in protein domains belonging to the NAD-binding Rossmann fold , as their co-factor binding sites are strikingly similar [11] . When entacapone and tolcapone were docked into 215 NAD-binding proteins from multiple species , the InhA enzyme from several different organisms , including M . tuberculosis , was consistently highly ranked ( see Tables S1 and S2 ) . Since InhA is the primary target of the anti-tubercular drugs isoniazid [17] and ethionamide [18] , entacapone and tolcapone can potentially inhibit the InhA ligand binding site directly . Indeed , alignment of the COMT and InhA binding sites by the SOIPPA algorithm revealed similarities in the positioning of both their co-factors and ligands ( Figure 1 ) . As shown in Figure 2 , the existing InhA inhibitor with the greatest 2D similarity to entacapone is 3- ( 6-aminopyridin-3-yl ) -N-methyl-N-[ ( 1-methyl-1H-indol-2-yl ) methyl]acrylamide ( AYM ) [23] ( Tanimoto coefficient = 0 . 155 ) , whereas the existing InhA inhibitor with the greatest 2D similarity to tolcapone is 3-[ ( acetyl-methyl-amino ) -methyl]-4-amino-N-methyl-N-[ ( 1-methyl-1H-indol-2-yl ) -methyl]-benzamide ( ZAM ) [23] ( Tanimoto coefficient = 0 . 173 ) ( see Table S3 ) . Neither of their p-values ( AYM; 0 . 065 and ZAM; 0 . 205 ) are significant at the 0 . 05 level , implying that none of the investigated InhA inhibitors exhibit significant molecular similarity to either entacapone or tolcapone . Therefore , it is unlikely that ligand-based screening methods would be able to identify entacapone and tolcapone as potential InhA inhibitors . Table 1 shows the predicted binding affinities of entacapone and tolcapone towards InhA . Since they fall within the range of binding affinities exhibited by the known InhA inhibitors , this not only provides a further implication of the cross-reactivity between InhA and COMT , but also suggests that entacapone and tolcapone are able to inhibit InhA directly . Unfortunately it is difficult to identify entacapone as a lead compound using conventional virtual screening because it is only ranked at 15 , 892 and 9 , 719 by eHiTs and Surflex among 20 , 000 randomly selected drug-like molecules , respectively . While advanced virtual screening techniques such as the relaxed complex scheme [24] , which combines a docking algorithm with molecular dynamics simulations , may improve the ranking of these potential lead compounds [25] , they demand significant computational resources . Thus , SOIPPA , a ligand binding site similarity based method , provides an efficient way of identifying potential drug-like leads that have well-established pharmacokinetics and pharmacodynamics properties . Although the 2D similarities between entacapone and existing InhA inhibitors have been shown to be statistically insignificant , entacapone shares a similar molecular size and common functional groups with several of the InhA inhibitors ( see Table 1 ) . For example , entacapone and five of the InhA inhibitors ( 468 , 566 , 641 , 665 and 774 ) all possess a single benzene and amide moiety . More importantly , the predicted binding poses of the benzene ring and the amide bond of entacapone are similar to those of the InhA inhibitors , with root mean square deviations ( RMSDs ) of as little as 2 . 87Å and 1 . 05Å , respectively ( in the case of 566 ) . Although tolcapone does not share the same amide moiety , the predicted binding pose of its benzene ring has an RMSD of only 1 . 01Å from that of 566 , further demonstrating the potential of these drugs as InhA inhibitors . Previous studies have highlighted the necessity of the interaction between the catechol oxygens of COMT inhibitors with an Mg2+ ion in the active site [26] . From the predicted binding poses of entacapone and tolcapone docked with InhA , three potential interaction sites of Mg2+ , including the Asp110 , Asp115 , and Glu210 residues of InhA , have been identified and are shown in Figure 3 . The closest residue , Glu210 , is positioned at a distance of 13 . 23Å from the nitrite group of entacapone , and at a distance of 11 . 54Å from the nitrite group of tolcapone ( see Figure S2 ) . Although it is possible that the conformation of the side chains may adjust to reduce this distance under in vitro or in vivo conditions , such a large distance may hinder the formation of a coordinate bond by the Mg2+ ion between the InhA active site and each of the two drugs . Consistent with the predicted binding pose , enzyme kinetic assays indicate that the addition of an Mg2+ ion has no effect on the inhibition of InhA by entacapone . This therefore provides us with opportunities to optimize entacapone and tolcapone , by reducing or removing their conjugation to the Mg2+ ion , so that they exhibit a weaker affinity for the original target COMT . The partition coefficient ( logP ) is the ratio of the concentrations of an unionized compound in the two phases of a mixture of octanol and water at equilibrium , whereas the distribution coefficient ( logD ) is the ratio of the sum of the concentrations of all forms of the compound ( both ionized and unionized ) in each of the two phases . Since logD is pH dependent , the pH at which it was measured is specified , as shown in Table 2 . Entacapone and tolcapone were discovered to have generally higher logP and logD values than most of the existing anti-tubercular drugs investigated . According to Lipinski's rule of five [27] , poor absorption or permeation is more likely for compounds with a logP value of greater than 5 . 0 . Such compounds are considered non-drug-like and are commonly filtered out in the early stages of drug discovery . However , entacapone and tolcapone are prescribed drugs and have been clinically tested with desired pharmacokinetics profiles . Indeed , entacapone can be rapidly absorbed with a Tmax of approximately one hour ( http://www . rxlist . com/comtan-drug . htm ) . The high logP values of entacapone and tolcapone are therefore acceptable regardless of Lipinski's rule . Moreover , they suggest that entacapone and tolcapone are more hydrophobic than existing drugs , and would therefore pass more easily through the M . tuberculosis cell envelope . Although entacapone demonstrates poor solubility , dissolution enhancers such as croscarmellose sodium are used in the formulation of Comtan [28] . This example well illustrates that drug repurposing would accelerate drug discovery and development by bypassing the time consuming steps of ADME/Tox evaluation and drug formulation . Entacapone and tolcapone were evaluated for their ability to inhibit growth of M . tuberculosis using a 96 well microplate assay . A 99 . 0% reduction in growth was observed with concentrations of between 62 . 5 µg/ml and 125 µg/ml for each drug . The sensitivity of M . tuberculosis to entacapone was confirmed by quantitative growth on agar plates containing known amounts of the drug . The minimum inhibitory concentration was between 62 . 5 µg/ml and 125 µg/ml ( Table 3 ) , therefore confirming the result from the microplate assay . These results support the computational predictions that entacapone and tolcapone have inhibitory activity against M . tuberculosis and may therefore be considered as lead compounds . The ability of entacapone to directly inhibit InhA was evaluated using enzyme kinetics . Due to the strong UV absorbance of entacapone over a wide range of 300–400 nm ( see Figure S4 ) , and the poor solubility of entacapone in water , the highest concentration of Comtan that could be used in the assays was 90 µg/ml ( the corresponding concentration of pure entacapone is 24 . 9 µg/ml ) , at which concentration InhA is approximately 47% inhibited . Fitting of the data in Table 4 to a dose response equation provided an IC50 value of 24±3 µg/ml ( 79±10 µM ) ( see Figure S3 ) . Since Mg2+ is critical for entacapone and tolcapone to inhibit COMT , assays were repeated in the presence of 5 mM Mg2+ to explore whether or not metal ion chelation could improve the affinity of the drug for InhA . However , the inclusion of Mg2+ in the assay had no effect on the IC50 value for enzyme inhibition . The progress curve analysis for the inhibition of InhA by Comtan shows that the UV absorbance is linearly time-dependent up to one hour ( see Figure S5 ) , indicating that entacapone is not a slow-onset inhibitor .
The continuing emergence of M . tuberculosis strains resistant to all existing , affordable drug treatments means that the development of novel , effective and inexpensive drugs is an urgent priority [3] . Our chemical systems biology approach to drug discovery revealed that Comtan , with the active component entacapone , shows potential for use as an anti-tubercular drug . Entacapone may adopt different inhibition mechanisms from the first- and second-line drugs that result in MDR and XDR M . tuberculosis strains . Moreover , it has an excellent safety profile with few side effects , and is commercially available . Therefore , entacapone can potentially be used as a lead compound to develop a new class of anti-tubercular drugs . By integrating techniques from ligand binding site similarity , small molecule similarity and protein-ligand docking , our chemical systems biology approach is able to model protein-ligand interaction networks on a proteome-wide scale . The systematic use of small molecules to probe biological systems will provide us with valuable clues as to the molecular basis of cellular functions , and at the same time it will shift the conventional one-target-one-drug discovery process to a new multi-target-multi-drug paradigm .
Previously , the SOIPPA algorithm had revealed a highly significant similarity ( p-value = 2 . 7e-5 ) between the NAD binding site of the Rossmann fold and SAM binding site of the SAM methyltransferases [11] . In order to identify similar ligand binding sites adjunct to the co-factor binding site , further docking studies were carried out on the ligand binding sites of proteins that bind NAD as a co-factor . Freely available docking software eHiTs 6 . 2 [41] and Surflex2 . 1 [42] were selected due to their relatively fast speed , high accuracy and ease of automation in large-scale docking studies . It is worth noting that when preparing PDB files for any of the docking studies , NAD and SAM co-factors were added as one of residues of the protein chain . 215 non-redundant proteins with NAD co-factors were downloaded from the RCSB Protein Data Bank ( PDB ) [40] , [43] . The ideal coordinates for entacapone and tolcapone were downloaded from DrugBank [44] . Both Surflex and eHiTs were used to dock both entacapone and tolcapone onto each of the 215 proteins , and the proteins that produced the highest docking scores were investigated further . Enoyl-acyl carrier protein reductases ( InhAs ) from several different organisms , including M . tuberculosis and Toxoplasma gondii , were identified as proteins to which entacapone and tolcapone showed favorable binding affinities ( see Tables S1 and S2 ) . However , the M . tuberculosis InhA ( PDB ID: 2H7M ) is the focus of this paper due to its importance as a drug target for the treatment of tuberculosis . Unfortunately , the human COMT protein that is the drug target of entacapone and tolcapone is absent from the RCSB PDB . The only COMT structure available at the time of writing is that from the brown rat , Rattus norvegicus . A standard protein BLAST [45] search revealed that human and rat COMT share 81% sequence identity without insertion or deletion . Moreover , their functional site residues were found to share 100% identity ( see Figure S1 ) . Therefore , rat COMT ( PDB ID: 2CL5 ) was used as an accurate representation of human COMT throughout this study . The SOIPPA algorithm was used to align the 2H7M structure with that of 2CL5 so that their respective NAD and SAM co-factor binding sites were aligned . The aligned proteins were then visualized using Accelrys DS Visualizer ( http://www . accelrys . com/products/downloads/ds_visualizer/ ) . The RCSB PDB was queried for proteins with sequence similarity to chain A of 2H7M , using an e-value cut off of 0 . 0001 . Between them , the resulting proteins bound a total of 22 different InhA inhibitors ( including the ligand of 2H7M ) . OpenBabel ( http://openbabel . org ) was used to calculate the 2D small molecule similarity between these InhA inhibitors and both entacapone and tolcapone . In order to create background distributions for comparison , all drug-like molecules were downloaded from the ZINC database [46] , and a subset of 20 , 000 molecules was extracted randomly . The 2D similarities of each of these molecules to both entacapone and tolcapone were calculated using OpenBabel , and density distributions of the scores were plotted using R 2 . 5 . 0 [47] . P-values corresponding to the 2D similarity scores of the 22 InhA inhibitors were subsequently calculated from both density distributions ( see Table S3 ) . Nine M . tuberculosis InhA structures ( PDB IDs: 1P44 , 1P45 , 2B36 , 2B37 , 2H7I , 2H7L , 2H7M , 2H7N and 2H7P ) were downloaded from the RCSB PDB . The ten inhibitors of these InhAs ( Ligands: Pyrrolidine carboxamide s3 , GEQ , TCL , 5PP , 8PS , 566 , 665 , 641 , 744 and 468 ) were downloaded from the RCSB PDB as ideal coordinates . In addition , the ideal coordinates of entacapone and tolcapone were generated using CORINA ( http://www . molecular-networks . com/online_demos/corina_demo . html ) . All twelve molecules were docked onto the nine InhA structures , as well as onto COMT using eHiTs and Surflex . The mean and standard deviation of the docking scores of each molecule with all ten of the InhAs were calculated , and the docking scores were tabulated for comparison . The predicted binding poses of entacapone and tolcapone with the various different M . tuberculosis InhAs were visualized and analyzed using Accelrys DS Visualizer . Distances between the nitrite groups of entacapone and tolcapone and the surrounding aspartic acid and glutamic acid residues of InhA were measured . RMSDs between the benzene rings and the amide bonds of entacapone and the native ligands were calculated , in addition to the RMSDs between the benzene rings of tolcapone and the native ligands . The ideal coordinates of entacapone , tolcapone , five first-line anti-tubercular drugs ( ethambutol , isoniazid , pyrazinamide , rifampicin and streptomycin ) and three second-line anti-tubercular drugs ( ciprofloxacin , moxifloxacin and aminosalicylic acid ) were downloaded from DrugBank [44] . Both ChemAxon's Calculator from Marvin Beans ( http://www . chemaxon . com/marvin ) and ChemSilico Predict ( http://www . chemsilico . com ) were used to calculate a ) the partition coefficient ( logP ) and b ) the distribution coefficient ( logD ) of all ten drug molecules . In order to prepare stock solutions of each drug , one tablet of Comtan ( Sandoz ) containing 200 mg of entacapone , and one tablet of Tasmar ( Valeant ) containing 100 mg of tolcapone were each ground to a fine powder and completely dissolved in dimethylsulfoxide ( DMSO ) . For the microplate assay , serial dilutions of each drug were made in Middlebrook 7H9 media supplemented with ADS ( albumin , dextrose , and saline ) and Tween 80 [48] in a volume of 100 µl . A culture of M . tuberculosis Erdman was grown to mid log in 7H9 plus ADS and Tween 80 , and adjusted to an optical density600 of 0 . 2 . 100 µl of bacteria was subsequently added to each well . The cultures were incubated for 14 days until the control wells containing only medium developed a dense layer of bacteria . Wells were visually scored for the amount of growth in comparison to the control wells . The dilution of drug that produced almost complete inhibition of growth was scored as MIC99 . The agar plate assay was carried out as previously described [48] using Middlebrook 7H9 plates supplemented with OADC and containing known amounts of entacapone . In order to determine the MIC99 , the number of bacteria that grew in the presence of each concentration of entacapone was compared with the number of bacteria that grew on the plate with no drug . Comtan ( entacapone ) tablets were ground into powder , and dissolved in DMSO . Kinetic assays using trans-2-dodecenoyl-Coenzyme A ( DD-CoA ) and wild-type InhA were performed as described previously [49] . Reactions were initiated by addition of InhA to solutions containing 250 µM NADH , 25 µM DD-CoA , 0 or 5 mM MgCl2 and inhibitor in 30 mM PIPES and 150 mM NaCl , pH 6 . 8 buffer . IC50 values were calculated by fitting the initial velocity data to equation 1; ( 1 ) where I is the inhibitor concentration and y is the percent activity . Data analysis was performed using Grafit 4 . 0 ( Erithacus Software Ltd . ) . The IC50 curve fitting is shown in the Figure S3 . A progress curve was calculated in order to study the slow-onset mechanism of inhibition of Comtan ( entacapone ) . InhA activity was monitored by adding the enzyme ( 10 nM ) to assay mixtures containing 8% V/V glycerol , 0 . 1 mg/ml BSA , 2% V/V DMSO , 300 µM DD CoA , 250 µM NADH , 200 µM NAD+ and inhibitors . Reactions were monitored until the progress curve became linear , therefore indicating that the steady-state had been reached . Subsequently , a low enzyme concentration and a high substrate concentration were used to ensure that the depletion of the substrates was minimal and would not affect the reaction rate , so that the progress curve in the absence of inhibitors was linear . Progress curve data were collected for up to 1 hour ( see Figure S5 ) . | The rise of multi-drug resistant ( MDR ) and extensively drug resistant ( XDR ) tuberculosis around the world , including in industrialized nations , poses a great threat to human health . This resistance highlights the need to develop new , effective and inexpensive anti-tubercular agents . Unfortunately , conventional approaches have yielded very few successes in the field of anti-infective drug discovery . It is a challenge to design drugs with both efficacy and safety . These challenges are reflected in the high costs involved in bringing new drugs to market . It has been estimated that the cost to launch a successful new drug is in excess of US$800 million . We have developed a novel computational strategy to systematically identify cross-reactivity between different drug target families . In this paper we demonstrate the strength of this approach through the discovery that existing commercially available drugs prescribed for the treatment of Parkinson's disease have the potential to treat MDR and XDR tuberculosis . The protocol described herein can be included in a drug discovery pipeline in an effort to accelerate the development of new drugs with reduced side effects . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
]
| [
"computational",
"biology/macromolecular",
"structure",
"analysis",
"pharmacology/drug",
"resistance",
"pharmacology/drug",
"development",
"biotechnology/small",
"molecule",
"chemistry",
"infectious",
"diseases/bacterial",
"infections",
"pharmacology/adverse",
"reactions",
"computational",
"biology/systems",
"biology"
]
| 2009 | Drug Discovery Using Chemical Systems Biology: Repositioning the Safe Medicine Comtan to Treat Multi-Drug and Extensively Drug Resistant Tuberculosis |
Skin sores caused by Group A streptococcus ( GAS ) infection are a major public health problem in remote Aboriginal communities . Skin sores are often associated with scabies , which is evident in scabies intervention programs where a significant reduction of skin sores is seen after focusing solely on scabies control . Our study quantifies the strength of association between skin sores and scabies among Aboriginal children from the East Arnhem region in the Northern Territory . Pre-existing datasets from three published studies , which were conducted as part of the East Arnhem Healthy Skin Project ( EAHSP ) , were analysed . Aboriginal children were followed from birth up to 4 . 5 years of age . Self-controlled case series design was used to determine the risks , within individuals , of developing skin sores when infected with scabies versus when there was no scabies infection . Participants were 11 . 9 times more likely to develop skin sores when infected with scabies compared with times when no scabies infection was evident ( Incidence Rate Ratio ( IRR ) 11 . 9; 95% CI 10 . 3–13 . 7; p<0 . 001 ) , and this was similar across the five Aboriginal communities . Children had lower risk of developing skin sores at age ≤1 year compared to at age >1 year ( IRR 0 . 8; 95% CI 0 . 7–0 . 9 ) . The association between scabies and skin sores is highly significant and indicates a causal relationship . The public health importance of scabies in northern Australia is underappreciated and a concerted approach is required to recognise and eliminate scabies as an important precursor of skin sores .
Remote Aboriginal communities in northern Australia have the world’s highest prevalence of skin sores with more than 80% of children affected by their first birthday[1] . Skin sores , also known as impetigo , are commonly caused by Group A Streptococcus ( GAS ) infections in these populations and can have serious sequelae such as invasive bacterial infection and post-streptococcal glomerulonephritis , which in turn increases the risk of chronic renal disease [2] . Acquisition of skin sores is influenced by other skin infections , particularly scabies . Scabies is endemic in remote northern Australia and found in up to 35% of children and 25% of adults in the region [3] . Scabies infection often leads to a secondary GAS infection of the skin and scabies control is considered a priority in measures aimed at reducing skin sores in the Aboriginal population [4–7] . Kearns et al . found that skin sores were seven times more likely to be concurrently diagnosed with scabies than when there was no scabies diagnosis[1] . The risk ratio was calculated based on the diagnosis of skin sore infection at a presentation of scabies compared to no diagnosis of scabies at the same presentation . However , their study focused on the concurrent risk between scabies and skin sores and further research is needed to understand the temporal relationship between the two conditions . Our study employed a self-controlled case series method to quantify the risk of scabies on skin sores using historical data from observational studies in Northern Territory . The study findings will contribute to the ongoing mathematical modelling research on understanding of GAS transmission and inform policy makers in prevention and control of skin sores and scabies in remote Aboriginal communities .
We analysed pre-existing datasets from three published studies , which were conducted as part of the East Arnhem Healthy Skin Project ( EAHSP ) overseen by the Menzies School of Health Research [1 , 8 , 9] . The EAHSP was a regional collaboration to reduce skin infections among children in five remote Aboriginal communities in the East Arnhem region in the Northern Territory . The three studies , namely Kearns et al . [1] , McMeniman et al . [9] , and Clucas et al . [8] , were separate but overlapping cohorts in the communities where EAHSP was conducted . These studies employed the same methodology to retrospectively review medical records from community health clinics in the region . Despite their similar research methodology , the three studies varied in inclusion criteria and duration of follow-up ( Table 1 ) . In Kearns et al . , eligible participants needed to have at least one clinic presentation in each quarter of their first year of life [1] . The duration of follow-up differed among the studies: participants were followed from birth to 1 year in Kearns et al . , to 2-years in McMeniman et al . , and up to 5-years of age in Clucas et al . One caveat of the Clucas et al . study is that although children born after 1 January 2001 were included , presentations were followed only from 1 January 2002 , and therefore not all participants were followed from birth ( 28% ) [8] . Complete descriptions of the research methods and study population for these studies are available elsewhere [1 , 8 , 9] . Data collected included date of presentation , child’s height and weight , any reason for presentation , antibiotic use , and referrals to hospital . Scabies were recorded as reasons for presentation if they were either noted specifically or with reference to scabies treatment given , while skin sores were recorded if there was any mention of skin sores or other presumed bacterial infections of the skin including boils , carbuncles , abscesses , ulcers and pustules ( Clucas ) . Multiple presentations on the same day were recorded as a single event . We merged the datasets from three published studies into a single composite line-listed dataset . All data analysed were de-identified . Ethics approval was obtained from the Human Research Ethics Committee of the Northern Territory Department of Health and Families and the Menzies School of Health Research ( Ethics approval 2015–2516 ) . Data on infection-free children i . e . , who developed neither scabies nor skin sores were not available , in the three pooled studies , to determine the difference between the exposed and non-exposed groups using cohort methods . Therefore , we used self-controlled case series ( SCCS ) method as an alternative to cohort studies by comparing the risks during different time periods within individuals[10] . The SCCS method compares the risks of developing events in the periods following exposure versus non-exposed periods within individuals . This method controls for fixed ( i . e . time-independent ) confounders such as gender and ethnicity , and only requires cases for analysis [11 , 12] . Our study included children from five Aboriginal communities who attended the community health clinics from January 2001 to February 2007 . We applied the following exclusions sequentially to our study cohort: The observation period ran from birth to the earliest of ( i ) the first time the observation period went for more than 120 days between clinic visits; ( ii ) the end of the study follow-up period; or ( iii ) the maximum age for their respective study ( Table 1 ) . The observation period was further divided in two age groups: ≤1 year and >1 year ( up to 4 years old ) to account the effect of age on both exposure and outcome . Scabies and skin sores episodes were identified from the clinic records . We assessed , within individuals , the risk of developing skin sores following scabies infection ( exposed period ) versus the risk of skin sores infection when there is no scabies ( non-exposed period ) . The period with the risk of developing skin sores was further subdivided into two segments: pre-exposure period and exposed period . The pre-exposure period is the period before the diagnosis of scabies ( the time taken from onset of scabies symptoms to the diagnosis of scabies by the clinician ) while the exposed period is the day of the diagnosis of scabies plus a specified period following the scabies diagnosis ( Fig 1 ) . The pre-exposure period is included to account for the presence of scabies prior to clinic attendance and any delay in the diagnosis of scabies . Our baseline assumption was that the pre-exposure period was seven days and exposed period was 14 days . These periods were defined based on the first diagnosis of scabies and any subsequent episodes that fell within the same risk period , were counted as one infection . Skin sores episodes were then mapped in relation to different time periods . A seven-day duration was chosen for the pre-exposure period based on the shortest time period reported in the literature [13 , 14] and the very frequent presentation of our participants for medical attendance ( on average every two weeks ) [1 , 8 , 9] . The natural history of skin sores is that an uncomplicated impetigo heals spontaneously within two weeks [15 , 16] . Detailed studies of the natural history of impetigo primarily caused by S . pyogenes found that the mean time to spontaneous healing was 12 . 6 days , with a range of 6–31 days [17] . These data and assumptions may not be generalisable to impetigo in non-endemic settings where Staphylococcus aureus is the primary pathogen . In our study , we assumed that repeat clinic presentations , positive for skin sores within a two-week period , were the same infectious episode and only the first episode was included in the analysis . We transformed the data into a time series format and applied deterministic imputation to substitute missing data as follows . The imputation was performed in a stepwise approach . Firstly , we found the first event ( skin sores or scabies ) in an individual and carried the value forward for the following 14 days . This step was repeated for any recurring events within the same individual . After that , any missing data were assumed as non-events . This was to ensure that skin sores and scabies events within 14 days were counted as one episode . Our primary analysis involved estimating the relative incidence rate of skin sores during exposed periods compared to the non-exposed period . Conditional Poisson regression was used to calculate incidence rate ratios ( IRRs ) and 95% confidence intervals ( CIs ) . Analyses were undertaken using Stata/IC version 14 ( StataCorp . College Station , TX , US ) . We converted the time periods from years to days for higher granularity and better accuracy . We undertook sensitivity analyses by increasing the exposed period from 14 to 21 and 28 days while keeping the pre-exposure interval consistent . We further analysed the incidence rate ratios within each community .
Compared to the non-exposed periods , the overall rate of skin sores was increased in both pre-exposure and exposed periods , with a significant increase in the exposed periods . Although the pre-exposure periods were associated with a higher prevalence of skin sores than baseline non-exposed periods ( IRR 1 . 3; 95% CI 0 . 8–1 . 9 ) , this finding was not statistically significant ( p = 0 . 296 ) . However , children were 11 . 9 times more likely to develop skin sores during the exposed periods compared with the baseline non-exposed periods ( IRR 11 . 9; 95% CI 10 . 3–13 . 7; p<0 . 001 ) ( Table 2 ) . When exposed to scabies , children had lower risk of developing skin sores at age ≤1 year compared to at age >1 year ( IRR 0 . 8; 95% CI 0 . 7–0 . 9 ) . Sensitivity analysis on the duration of the exposure period ( 14 days , 21 days and 28 days ) showed that increasing the length of the exposed period resulted in a progressive decrease in the incidence rate ratio for skin sores associated with scabies , compared with the baseline ( Table 2 ) . However , the study outcomes for pre-exposure periods and age groups did not change significantly ( Table 2 ) . Community stratified rates in five communities ( using the primary analysis assumption of 14-day exposure period ) showed similar results to the overall rates in both pre-exposure and exposed periods , with overlapping confidence intervals ( Table 3 ) .
Our study is the first to use the self-controlled case series method to investigate the association between scabies and skin sores . We found that , when infected with scabies , children were 12 times more likely to develop skin sores than in the absence of scabies infestation . Our findings reiterate the extreme burden of skin disease in remote Aboriginal and Torres Strait Islander communities , with up to 75% of children in these communities having a skin sore by their first birthday . This force of infection is compatible with data from low and middle-income regions such as India and those in the South Pacific where the prevalence of skin sores among preschool children and adolescents ranges from 42% to 70% [13] . The risk estimate from our study is substantially higher than previously reported relative risks of scabies and skin sores co-infections , which ranged from 2 . 4 to 7 . 0 [1 , 3 , 13 , 18] . There are two possible explanations . Firstly , earlier studies analysed risk based on concurrent presentation with scabies and skin sores—our study extended the hypothesised association to a temporal relationship between diagnosed scabies and skin sores , within a defined risk window . Secondly , the SCCS method has the advantage of controlling implicitly for fixed confounders , which can affect case-control and cohort studies , and would be anticipated to report a more faithful estimate of true risk [19] . Whilst our study cannot establish causality definitively , it has a number of attributes supporting a causal relationship between scabies and skin sores . These include the significant association and the temporality between the two conditions and consistency with earlier studies . These results are complemented by the sensitivity analyses on the duration of the post diagnosis scabies exposure period . We found that lengthening the exposure window from 14 to 21 days and then to 28 days was associated with a decline in the incidence rate ratio . This strengthens inference on the purported causal relationship between scabies and skin sores , i . e . , if scabies truly increases the risk of skin sores , it is logical that the risk of this outcome will occur within a time frame more narrowly associated with the diagnosis ( i . e . 14 days rather than 28 days ) . One of the strengths of our study is inclusion of pre-exposure period in the analysis so as to reduce the inflation of relative risks in the exposed period . We observed that the risk of developing skin sores in pre-exposure periods is much lower than that in exposed periods . However , the results for pre-exposure periods are not statistically significant , and furthermore , these findings are probably due to clinic nonattendance rather than actual absence of infections . We found that children were less likely to develop skin sores in their infanthood than when they were between one and four years of age . This is consistent across different durations of scabies infection . One possibility is that infants are less likely to scratch and develop excoriation of the skin when infected with scabies than older children , thereby reducing the incidence of skin sores . However , it is important to note that the overall association between skin sores and scabies remains significant regardless of age . Our finding is of a link at the individual level between scabies and skin sore risk . Studies of skin sores in the East Arnhem region have identified GAS as the dominant pathogen in around 80% of cases [2] , and clearance of GAS has been identified as the only independent predictor of complete resolution of skin sores [20] . It is therefore likely that the observed association between scabies and skin sores reflects a heightened risk of streptococcal infection in the presence of scabies , but we can neither confirm nor refute a similar association with staphylococcal infection on the basis of the evidence here presented . The findings are unlikely to be generalisable to non-endemic settings for impetigo where S . aureus is usually the primary pathogen . Our study limitations include that , in the SCCS design , occurrence of an event should not affect subsequent exposures . Although research is limited in the area of scabies and skin sores , to date , there is no evidence suggesting the impact of skin sores on the risk of developing scabies . Therefore , we assumed that a skin sore infection was unlikely to affect subsequent scabies exposure . However , our exposures and events are defined by diagnosed cases and may be influenced by clinic attendances . For example , if a diagnosis of skin sores prompted a participant to return to the clinic for clearance check-up , these follow-up visits may increase the chance of scabies being detected incidentally . Secondly , our study has likely underestimated the incidence of scabies as we only included the children who presented to the clinics . Furthermore , we did not control for any potential time varying confounders , which can cause delay in diagnosis such as stigma associated with scabies , restricted access to health care and chronic steroid use causing masked presentation[21–23] . Thirdly , our study relies on the documentation and clinical diagnosis of the clinicians . While there is imprecision involved in relying on syndromic reporting of skin sores and clinically observed scabies identified from an historical clinic record review , this level of diagnostic uncertainty reflects the reality of clinical practice in a remote setting . We anticipate that the high prevalence of both conditions in the communities studied should increase clinical experience and the consistency and reliability of observer reporting . Indeed , the involved communities and associated healthcare clinics have had a long standing interest in skin health related research , and thus have well engaged clinicians around the diagnosis and management of impetigo and scabies . Moreover , our use of the self-controlled case series method accounts for any variability between participants . Lastly , whilst the SCCS method provides a robust study design , we may not be able to generalise the findings conclusively . Our findings are based on a specific subgroup of Aboriginal children and may not be representative of the wider Australian population or older age groups nor perhaps of Aboriginal children living in other remote communities . Furthermore , SCCS method was first developed for rare events and historically used in vaccine trials , yet we assessed two common conditions in Aboriginal communities . However , others have utilised the SCCS method for common events such as drug safety and have produced findings consistent with those of previous studies , which suggests the SCCS method was appropriate for our study[24 , 25] . The self-controlled case series method we have used is potentially transferrable to other settings where data have been collected that do not invalidate the necessary conditions . By investigating associations within an individual , the self-controlled case series method implicitly controls for contextual factors that may influence the risk of either infection , making findings more comparable across situations and settings . As regards to public health implications , our findings robustly support existing inference on the contribution of scabies to skin sore risk , particularly in the very young [13 , 14] . Accordingly , we endorse holistic healthy skin strategies that raise awareness of the need for prevention , early presentation and multi-pronged treatment strategies to reduce the overall burden and the long-term sequelae of skin disease . In conclusion , our study demonstrated that the association between scabies and skin sores is significant , more so than previously reported in standard cohort and case- controlled studies . A concerted approach is needed in implementing scabies elimination programs to prevent skin sores , particularly due to GAS , and their devastating complications in remote Aboriginal communities . | Skin sores , also known as impetigo , are highly contagious bacterial skin infections , which are found commonly in school children and occasionally in adults . Skin sores are prevalent in disadvantaged or resource-poor settings . In Australia , about two thirds of Aboriginal children suffer from skin sores by their first birthday . If untreated or treated poorly , skin sores can eventually cause heart and kidney problems . It is also believed that scabies , another common skin infection in Aboriginal children , can increase the risk of developing skin sores by allowing the bacteria to enter the skin more easily through breaks in the skin . Our research explored the following: if scabies is a risk factor for skin sores then what is the strength of the association between the two conditions . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
]
| [
"dermatology",
"medicine",
"and",
"health",
"sciences",
"clinical",
"research",
"design",
"tropical",
"diseases",
"geographical",
"locations",
"australia",
"group",
"a",
"streptococcal",
"infection",
"parasitic",
"diseases",
"research",
"design",
"bacterial",
"diseases",
"age",
"groups",
"skin",
"infections",
"ectoparasitic",
"infections",
"case",
"series",
"sexually",
"transmitted",
"diseases",
"neglected",
"tropical",
"diseases",
"skin",
"diseases",
"research",
"and",
"analysis",
"methods",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"scabies",
"people",
"and",
"places",
"oceania",
"population",
"groupings"
]
| 2018 | Scabies and risk of skin sores in remote Australian Aboriginal communities: A self-controlled case series study |
Recent data suggest that Nef-mediated downmodulation of TCR-CD3 may protect SIVsmm-infected sooty mangabeys ( SMs ) against the loss of CD4+ T cells . However , the mechanisms underlying this protective effect remain unclear . To further assess the role of Nef in nonpathogenic SIV infection , we cloned nef alleles from 11 SIVsmm-infected SMs with high ( >500 ) and 15 animals with low ( <500 ) CD4+ T-cells/µl in bulk into proviral HIV-1 IRES/eGFP constructs and analyzed their effects on the phenotype , activation , and apoptosis of primary T cells . We found that not only efficient Nef-mediated downmodulation of TCR-CD3 but also of MHC-I correlated with preserved CD4+ T cell counts , as well as with high numbers of Ki67+CD4+ and CD8+CD28+ T cells and reduced CD95 expression by CD4+ T cells . Moreover , effective MHC-I downregulation correlated with low proportions of effector and high percentages of naïve and memory CD8+ T cells . We found that T cells infected with viruses expressing Nef alleles from the CD4low SM group expressed significantly higher levels of the CD69 , interleukin ( IL ) -2 and programmed death ( PD ) -1 receptors than those expressing Nefs from the CD4high group . SIVsmm Nef alleles that were less active in downmodulating TCR-CD3 were also less potent in suppressing the activation of virally infected T cells and subsequent cell death . However , only nef alleles from a single animal with very low CD4+ T cell counts rendered T cells hyper-responsive to activation , similar to those of HIV-1 . Our data suggest that Nef may protect the natural hosts of SIV against the loss of CD4+ T cells by at least two mechanisms: ( i ) downmodulation of TCR-CD3 to prevent activation-induced cell death and to suppress the induction of PD-1 that may impair T cell function and survival , and ( ii ) downmodulation of MHC-I to reduce CTL lysis of virally infected CD4+ T cells and/or bystander CD8+ T cell activation .
Although all primate lentiviruses are referred to as human and simian “immunodeficiency” viruses ( HIV and SIV , respectively ) , SIVs do generally not cause disease in their natural monkey hosts [1] , [2] . This is in contrast to HIV-1-infected humans and SIVmac-infected macaques , who develop immunodeficiency and AIDS in the absence of antiretroviral therapy [3] . The exact mechanisms underlying nonpathogenic SIV infection are still unclear . Viral loads in naturally infected sooty mangabeys and African green monkeys are just as high as in HIV-1-infected humans and SIVmac-infected macaques [4]–[6] . Furthermore , SIVsmm and SIVagm replicate efficiently in lymphoid tissues and cause massive destruction of CD4+ memory T cells in the gut during acute infection [7] , [8] . However , a consistent difference between pathogenic and nonpathogenic primate lentiviral infections is that high levels of chronic immune activation associated with accelerated T cell turnover and apoptosis are observed in HIV-1-infected humans and SIVmac-infected macaques but not in naturally SIV-infected primates [1] , [2] . Increased immune activation is also a key predictor of progression to AIDS in HIV-1-infected humans [9] , [10] . It is thus believed that infection-associated chronic immune activation and T cell apoptosis drive the progressive destruction of the human immune system . One viral factor that modulates viral fitness , persistence and pathogenicity , is the viral Nef protein . Initially , Nef appeared to represent a potent “virulence” factor because disrupted nef genes were associated with very low viral loads and an attenuated clinical course both in HIV-1 and SIVmac infections [11]–[13] . Subsequently , a surprising number of different Nef functions have been described , including modulation of cell surface expression of CD4 , CD28 , class I MHC and the invariant chain associated with immature class II MHC molecules , as well as enhancement of viral infectivity and replication [14]–[21] . In pathogenic HIV-1 and SIVmac infections , the combination of these Nef functions promotes viral persistance and accelerates disease progression by facilitating viral immune evasion and by enhancing viral spread [22]–[27] . However , high viral loads are not associated with the development of disease in natural SIV infection [1] , [2] . Thus , Nef does not act as a pathogenesis factor in natural SIV infections . This may be due to the fact that most primate lentiviral Nef proteins efficiently downmodulate CD3 , a key component of the T cell receptor ( TCR ) , from the surface of virally infected T cells , while Nef proteins of HIV-1 and its immediate SIV precursors fail to perform this function [28] . It is controversial whether the HIV-1 Nef enhances or reduces the responsiveness of virally infected T cells to activation [28]–[36] . However , it is clear that nef alleles from primate lentiviruses which are capable of downmodulating TCR-CD3 suppress T cell activation and apoptosis substantially more efficiently than those that can not perform this function [28] . Recently , it has been shown that a small subset ( 10–15% ) of naturally infected SMs experience a significant loss of CD4+ T cells during the course of their infection [37] , [38] . Following up on this observation , we examined the Nef functions of the viruses infecting these animals . Interestingly , we found that efficient Nef-mediated downmodulation of TCR-CD3 was preserved only in those animals with high CD4+ T cells , but not in those with low CD4+ T cells [28] . This study provided a first indication that Nef may exert a protective role in natural SIV infections; however , it had limitations . First , viruses from only few SMs with low CD4+ T cells counts were available for study . Second , Nef functions were only analyzed in Jurkat T cells following transient transfection with vectors coexpressing Nef and eGFP . Finally , only few Nef activities , i . e . downmodulation of CD4 , CD3 , CD28 and MHC-I , were investigated . In the present study , we examined the effects of nef alleles from 11 SMs with CD4+ T cell counts >500 cells/µl ( CD4high Nef ) and 15 SMs with CD4+ T cell counts <500 CD4+ ( CD4low Nef ) on the surface expression of various cellular receptors , T cell activation and apoptosis of virally infected cells . Our results show that not only downmodulation of TCR-CD3 but also of MHC-I correlates with the preservation of CD4+ T cells in infected managbeys in vivo . Thus , primate lentiviral Nef proteins may exert a protective effect in natural SIV infection through two different yet complementary pathways by preventing activation-induced cell death of infected CD4+ T cells as well as CD8+ T cell activation and CTL lysis .
To study the function of primary SIVsmm nef alleles from SMs with different CD4+ T cell counts , we cloned nef alleles from 15 animals with low ( <500/µl ) and 11 animals with high ( >500/µl ) CD4+ T cell counts ( CD4low and CD4high Nefs , respectively ) into an HIV-1 NL4-3-based IRES-eGFP proviral vector coexpressing Nef and eGFP from bicistronic RNAs [28] , [39] . Nef alleles were amplified from the plasma of 22 naturally and four experimentally SIVsmm-infected SMs housed at the Yerkes National Primate Research Center ( Atlanta , GA ) and cloned in bulk into the proviral vector ( Figure S1 ) . Nef alleles from 19 of these 26 SMs were previously cloned into the bicistronic pCGCG expression plasmid and analyzed in Jurkat T cells by transient transfection [28] . The virological and immunological characteristics of this SM colony , particularly the correlates of CD4+ T cell depletion have recently been described [37] . The 15 “CD4low” animals include 12 of 14 previously reported naturally SIVsmm-infected SMs with CD4+ cell counts <500/µl [37] . The remaining three CD4low animals were infected by i . v . injection of plasma from a naturally SIVsmm-infected SM [38] . Phylogenetic analyses of amplified nef sequences verified that all proviral plasmid preparations contained nef alleles from different animals . Moreover , to ensure appropriate representation of the nef alleles present in vivo viral stocks were derived from ≥50 independent transformants ( data not shown ) . Alignment of the deduced amino acid sequences showed that several domains and putative protein interaction sites previously described to be relevant for HIV-1 Nef function [40] were conserved in SIVsmm Nefs , including the N-terminal myristoylation signal , the acidic region , a diarginine motif , a C-proximal adaptor-protein interaction site and a diacidic putative V1H binding site ( Figure 1 ) . In comparison , the P ( xxP ) 3 motif , which is highly conserved in HIV-1 Nef's and interacts with cellular kinases [40] , is not conserved in SIVsmm Nef alleles , which usually contain only two proline residues at this location ( Figure 1 ) . Inspection of the Nef sequences from CD4high and CD4low groups of SMs failed to reveal obvious inactivating mutations , such as large deletions or premature stop codons . However , Nef sequences from the CD4low group were more likely to contain changes in the N-proximal region ( residues 31 to 40 ) , the acidic motif and the region downstream of the proline-rich region . Moreover , changes in S63 or E64 were observed in 11 of 15 CD4low but only in 1 of 11 CD4high Nef alleles ( Figure 1 ) . Finally , substitutions of S138 , E139 , E191 , Q221 and T225 just upstream of the “diarginine motif” , the “dileucine”-based sorting signal ( ExxxLV ) [41] , and Y226 were substantially more frequent in Nef alleles from the CD4low group of SMs . Y226 corresponds to Y223 in the SIVmac239 Nef and is required for MHC-I downmodulation but not for other Nef activities [24] , [42] . Altogether , the sequence comparisons indicated that SIVsmm nef genes from SMs with both high and low CD4+ T cell counts encoded full-length Nef proteins but frequently differed in amino acid residues flanking known functional motifs . As reported previously [28] , the proviral NL4-3 Nef/eGFP constructs are isogenic except for their nef coding sequences and have the advantage that Nef expression is mediated by the viral LTR promoter and via the normal splice sites . Moreover , pseudotyping with the VSV-G envelope protein bypasses the effect of Nef on virion infectivity [43] and hence allows to transduce primary T cells with comparable efficiency independently of their nef coding regions . Finally , since infected cells coexpress Nef and eGFP from single bicistronic RNAs , Nef expression can be correlated with receptor expression levels , activation status and apoptosis [28] , [39] . To compare the potency of primary SIVsmm nef alleles from SMs with differential CD4+ T cell counts in downmodulating CD3 , CD4 , CD28 , MHC-I and CXCR4 , we transduced Jurkat T cells with the proviral constructs and analyzed them by flow cytometry ( examples shown in Figure S2 ) . Consistent with earlier results [28] , we found that CD4high nef alleles were significantly more active in TCR-CD3 downmodulation than those from the CD4low animals ( 7 . 2±0 . 8 , n = 11 vs 4 . 2±0 . 5 , n = 15; P = 0 . 0033; numbers give average n-fold downmodulation ±SD ) ( Figure 2 ) . However , the present analysis of a larger number of nef alleles also revealed significant differences in MHC-I downmodulation between the CD4high and CD4low groups ( 3 . 0±0 . 1 , n = 11 vs 2 . 3±0 . 1 , n = 15; P = 0 . 0023 ) . Exclusion of nef genes from four experimentally infected SMs did not change these results: the differences in CD3 and MHC-I modulation remained significant ( P = 0 . 027 and P = 0 . 011 , respectively ) . In comparison , the two sets of nef alleles did not differ significantly in their abilities to downmodulate CD4 , CD28 and CXCR4 from the surface of infected Jurkat cells . Notably , the detectable effect of Nef on CD4 was weak because the proviral constructs express Vpu and Env which also downmodulate this receptor [44] , [45] . Since SIVsmm specific Nef antibodies are not available , the contribution of Nef expression levels to the observed functional differences could not be assessed . However , the fact that the majority of nef alleles were functionally active at least in some assays implies that they are efficiently expressed . Next , we examined Nef function in human PBMCs infected with eGFP-expressing reporter viruses ( data summarized in Table S1 ) . Importantly , these experiments were performed in both the presence and absence of CD8+ T cells ( examples shown in Figure 2A and S2 ) to ( i ) maintain some of the complexity that exists in vivo and ( ii ) to ensure that transduction of CD8+ T cells ( that are usually not infected with SIV and HIV ) does not impact the results . Flow cytometric analyses of virally infected PBMCs confirmed the correlation between efficient TCR-CD3 ( 4 . 0±0 . 3 vs 2 . 4±0 . 2; P = 0 . 0002 ) and MHC-I ( 2 . 8±0 . 1 vs 2 . 2±0 . 1; P = 0 . 0004 ) downmodulation and preserved CD4+ T cell counts ( Figure 2B , middle ) . On average , PBMCs infected with reporter viruses containing CD4high nef alleles also expressed lower levels of the CD4 molecule , but this difference was less pronounced . In comparison , the effect of all SIVsmm nef alleles on CD28 and CXCR4 surface expression on PBMCs was weak and did not differ significantly between the CD4high and CD4low groups . CD8-depleted PBMC cultures yielded identical results ( Figure 2B , right ) . Finally , correlation analyses demonstrated that all nef alleles that efficiently downmodulated CD4 , TCR-CD3 and MHC-I in Jurkat cells were also highly active in PBMCs ( Figure S3 ) . To further evaluate the role of TCR-CD3 and/or MHC-I downmodulation by Nef in the maintenance of stable CD4+ T cell numbers , we performed correlation analyses . As shown in Figures 3A and S4A , the efficiency of these Nef functions correlated with the numbers of CD4+ T cells in SIVsmm-infected SMs in vivo . The correlations between both absolute and relative CD4+ T cell counts and the potency of Nef-mediated downmodulation of TCR-CD3 or MHC-I by Nef remained significant when only the 22 naturally SIVsmm-infected animals were included in the analyses ( Figures 3A , 3B and S4A , S4B ) . The strongest correlation between high numbers or percentages of CD4+ T cells and Nef function was obtained for the sum of TCR-CD3 and MHC-I downmodulation ( Figure S4C , S4D ) , suggesting that both functions synergize to protect SIVsmm-infected SMs against the loss of CD4+ T cells . We next examined whether these specific Nef functions correlated with additional virological or immunological parameters determined for this natural SIV infection cohort . As shown in Figures 3C and S5A , we found that the efficiency of Nef-mediated downmodulation of CD3 correlated with higher numbers of proliferating CD4+Ki67+ cells in SIVsmm-infected SMs . The same trend was observed for modulation of MHC-I by Nef but failed to reach significance ( Figures 3C , S5A ) . In addition , we observed a significant inverse correlation between the efficiency of both TCR-CD3 and MHC-I downmodulation and the levels of CD95 expression by CD4+ T cells ( Figures 3D , S5B ) , suggesting that the levels of CD95-mediated apoptosis of CD4+ T cells as well as their effect on CD8+ bystander T cells [46] may be reduced in SMs infected with SIVsmm Nef variants that effectively downmodulate CD3 and MHC-I . We also found that efficient CD3 and MHC-I downmodulation correlated with higher numbers of CD8+CD28+ T cells ( Figures 3E , S5C ) . Notably , it has been shown previously that decreases in the absolute numbers of CD8+CD28+ T cells are predictive for progression to AIDS in HIV-1 infected individuals [47] and associated with declining CD4+ T cell counts in SIV-infected SMs [37] . Finally , we observed that the number of naïve CD4+ T cells correlates with the efficiency of MHC-I but not TCR-CD3 downmodulation by Nef ( Figures 3F , S5D ) . In contrast , we did not find significant correlations between Nef function and other immological parameters examined , such as the absolute numbers of CD8+ T cells and the levels of proliferating ( i . e . Ki67+ ) CD8+ T cells or of activated ( i . e . CD69+ or HLA DR+ ) , effector ( i . e . CD127− ) and regulatory CD25+CD4+ T cells . To identify further possible consequences of TCR-CD3 and MHC-I downmodulation , we compared the immunological and virological features of SMs naturally infected with SIVsmm expressing nef alleles showing the highest ( n = 6 ) or lowest ( n = 6 ) activities in these two functions , i . e . >4 . 0- vs <2 . 1-fold and >2 . 8- vs <2 . 2-fold downmodulation in PBMCs . While the results of subgroup analyses need to be interpreted with caution they are a meaningful approach to generate and assess new hypotheses [48] . This analysis confirmed that lack of these Nef functions correlates with reduced numbers and/or percentages of CD4+ T cells , CD4+Ki67+ and CD8+CD28+ T cells in SIVsmm-infected SMs ( Figure 4A–D ) but not with significant differences in viral loads ( Figure 4E ) . As expected from the results of the correlation analyses ( Figure 3F ) , efficient MHC-I but not TCR-CD3 downmodulation by Nef was associated with a reduced number of naïve CD4+ T cells in SIVsmm-infected SMs ( Figure 4F ) . Furthermore , this comparative analysis showed that the percentage of effector CD8+ T cells is significantly lower and that of naïve and memory CD8+ T cells significantly higher in animals infected with SIVsmm strains that effectively downregulated MHC-I ( Figure 4G–I ) . We also observed a trend for higher percentages of CD8+CCR7+ T cells ( Figure 4J ) . Taken together , these data demonstrate that effective TCR-CD3 and MHC-I downmodulation by Nef is a strong correlate of preserved CD4+ T cells counts in natural SIVsmm infection of SMs and further support that inefficient MHC-I antigen presentation by virally infected cells affects the CD8+ T cell response . We have previously shown that TCR-CD3 downmodulation by Nef blocks the responsiveness of virally infected T cells to activation [28] . To assess whether the relatively subtle functional differences of primary SIVsmm nef alleles in TCR-CD3 modulation are also associated with differential responsiveness of infected T cells to stimulation we first determined the surface expression levels of CD69 and of the IL-2 receptor-α chain ( IL-2Rα ) or CD25 , well known markers for early and late T cell activation , in HIV-1-infected PHA-treated regular and CD8-depleted PBMC cultures . These analyses showed that T cells infected with viruses expressing CD4high Nef alleles expressed significantly lower levels of activation markers than those infected with CD4low Nef constructs ( Figures 5A , S6A ) . In agreement with our previous data [28] , suppression of CD69 and IL-2R expression correlated with effective downmodulation of TCR-CD3 ( Figures 5B , S6B , S6C ) and reduced levels of programmed death ( Figure 5C ) . Furthermore , Nef-mediated downmodulation of CD3 ( Figure 5D ) but not that of MHC-I ( Figure 5E ) , correlated with reduced percentages of apoptotic cells . Although downmodulation of CD3 and MHC-I are mediated by different domains in Nef [24] , [26] , [49]–[51] , both functions correlated with one another ( Figure 5F ) , suggesting that both may be coselected in SIVsmm-infected SMs . Recent findings suggest that high levels of programmed death ( PD ) -1 receptor expression contribute to the exhaustion of HIV-1-specific CD8+ T cells because blocking PD-1-PD-1 ligand interaction enhanced their ability to proliferate and to produce cytokines in HIV-1 infected individuals [52] , [53] . PD-1 is also expressed on activated CD4+ T cells [54] , [55] and its blockade partly restored CD4 proliferative responses in AIDS patients [52] . It has been shown that PD-1 is inducibly expressed on activated T cells [56] . Since CD3 downmodulation prevents T cell activation , we examined whether this Nef function affects the induction of PD-1 expression by virally infected T cells . We found that PHA stimulation induced a 2 . 5- to 4-fold increase in PD-1 expression levels ( example shown in Figure 6A ) . Infection with HIV-1 constructs expressing no Nef or Nef alleles that do not downmodulate TCR-CD3 resulted in slightly increased levels of PD-1 expression . In strict contrast , those that downregulated TCR-CD3 generally suppressed the induction of this receptor ( Figure 6A ) . Further analyses demonstrated that Nef alleles from the CD4high group of SMs suppressed the induction of PD-1 expression significantly more efficiently than those derived from the CD4low group of animals ( Figure 6B ) . Since both represent T cell activation markers , we observed a highly significant correlation between the levels of IL-2R and PD-1 expression on HIV-1 infected PBMCs ( Figure 6C ) . Suppression of the induction of PD-1 expression correlated with the efficiency of TCR-CD3 downmodulation ( Figure 6D ) . The correlation between the ability of Nef to downmodulate CD3 and to suppress the induction of CD69 , PD-1 and IL-2R was confirmed using CD8-depleted PBMCs ( Figure S6 ) . Thus , the ability of Nef to modulate TCR-CD3 may impact the functionality , proliferation and survival of T cells by affecting the expression levels of PD-1 on virally infected CD4+ cells . The nuclear factor of activated T cells ( NFAT ) regulates transcription of IL-2 gene expression , a hallmark of T cell activation . It has been established that NFAT activation is suppressed by Nef alleles that downmodulate TCR-CD3 , e . g . SIVsmm and HIV-2 Nefs , and enhanced by Nef alleles that do not modulate this receptor , such as those of HIV-1 [28] , [31] . To assess whether nef alleles from the CD4high and CD4low groups of SIVsmm-infected SMs differentially affect NFAT induction in virally infected cells , we transduced Jurkat T cells stably transfected with the luciferase gene under the control of an NFAT-dependent promoter [31] , with the proviral HIV-1 eGFP/Nef constructs and examined their responsiveness to activation . T cells infected with nef defective HIV-1 showed about 5-fold enhanced levels of NFAT activity after PHA stimulation compared to mock infected cells ( Figure 7A ) . This increase was usually further enhanced by 2- to 3-fold by expression of HIV-1 and SIVcpz Nef alleles but blocked by the control SIVmac and HIV-2 Nefs ( examples shown in Figure 7A ) . SIVsmm Nef alleles from 25 of the 26 infected SMs also suppressed NFAT induction , albeit with differential efficiency . Interestingly , nef alleles from animal FYb , that showed the most severe CD4+ T cell depletion ( 3/µl ) of the entire colony of 110 naturally SIVsmm-infected SMs [37] , rendered the Jurkat T cells hyper-responsive to activation - just like those of HIV-1 ( Figure 7A ) . This unusual property of bulk FYb Nef alleles , which were generally poorly active in other functions including downmodulation of CD3 and MHC-I , was confirmed using 18 HIV-1 IRES/eGFP constructs expressing individual FYb derived Nef proteins ( data not shown ) . Compared to FYb , the NFAT-dependent luciferase activities were 4- to 13-fold lower in Jurkat cells infected with viral constructs expressing Nef alleles derived from the remaining 25 SMs . However , CD4low Nef alleles were less potent in suppressing NFAT induction than those from the CD4high group ( Figure 7B ) . The levels of NFAT activation in HIV-1-infected T cells inversely correlated with the CD4+ T cell counts of the respective SIVsmm-infected SMs in vivo , although this correlation was not very stringent ( Figure 7C ) . In comparison , the ability of these 25 nef alleles to suppress NFAT induction upon PHA stimulation did not correlate with their potency in downmodulating TCR-CD3 ( Figure 7D ) . This is most likely due to the fact , that the functional differences in modulation of TCR-CD3 between these SIVsmm nef alleles are much more subtle than those between Group 1 and 2 nef alleles [28] , and that other Nef functions , such as downmodulation of CD4 , CD28 and CXCR4 , also affect the responsiveness of virally infected T cells to activation . Nevertheless , these results further support that high responsiveness of infected T cells to activation accelerates the loss of CD4+ T cells in naturally SIV-infected SMs . In addition to facilitating viral immune evasion , Nef has been reported to also enhance virion infectivity [14] , [15] , [17] , [18] . To examine whether CD4+ T cell loss in SIVsmm-infected SMs is associated with differences in this Nef function , we exposed the HeLa-CD4/LTR-lacZ indicator cell line , P4-CCR5 , which expresses CD4 and both major entry cofactors CXCR4 and CCR5 [57] , [58] , to 293T cell-derived virus stocks and determined the β-galactosidase activities after two days . Most bulk SIVsmm nef alleles enhanced virion infectivity , albeit with varying efficiencies ( Figure 8A ) . Analysis of the same reporter viruses using the TZM-bl indicator cell line , previously designated JC53-bl [59] , [60] , yielded the same results ( Figure 8B ) . On average , nef genes from SIVsmm-infected SMs with high CD4+ T cell counts were slightly more active in enhancing virion infectivity than those from animals with low numbers of CD4+ T cells ( Figure 8C ) . However , this difference was mainly due to the unusually high activity of nef alleles from a single animal ( FAi ) and not significant . Moreover , the potency of Nef in enhancing virion infectivity did not correlate with the CD4+ T cell counts in SIVsmm-infected SMs ( Figure 8D ) . To analyze the effect of sequence variations between the CD4high and CD4low groups of SMs on Nef function we synthesized the consensus nef sequences derived from the six animals with the highest ( SMhi ) and the six SMs with the lowest ( SMlow ) numbers of CD4+ T cells , which differed in a total of 20 amino acid residues ( Figure S7A ) . We found that the SMhi Nef allele was moderately more active in downmodulating CD3 and MHC-I and in suppressing T cell activation and programmed death than the SMlow consensus Nef ( Table S2 ) . Thus , the SMhi and SMlow Nef alleles exhibited functional differences reflecting those observed for the primary nef alleles derived from SMs with high and low CD4+ T cell counts . However , both consensus Nefs were more active than most primary Nef alleles . This phenotype is similar to that of consensus Nef alleles from HIV-1-infected individuals with progressive and nonprogressive infection , which also displayed unusually high activity and only subtle functional differences [23] . The reason for this phenotype is most likely that these consensus Nefs contain optimal amino acid residues at most positions , although the possibility that they may be expressed at particularly high levels cannot be exluded . To further assess which amino acid variations modulate the ability of Nef to downregulate CD3 and MHC-I we analyzed about 50 individual primary SIVsmm nef alleles . The results confirmed the data obtained from the analyses of Nef function in bulk . For example , the FCs clone 1 Nef allele derived from a CD4high SM efficiently downmodulated CD3 and suppressed T cell activation , whereas the FYb clone 17 Nef from a CD4low animal was poorly active ( Table S2 ) . To map residues underlying these functional differences , we replaced eight amino acid residues in the FCs clone 1 Nef with the corresponding residues in FYb and introduced the reverse changes in the FYb clone 17 Nef ( Figure S7B ) . Our data showed that these changes impaired the activity of the FCs clone 1 Nef in receptor downmodulation but rendered it active in causing hyper-responsiveness of T cells to activation–similarly to those of FYb . In contrast , the reverse changes partially restored the ability of the FYb clone 17 Nef to downmodulate CD3 and to suppress T cell activation and programmed death ( Table S2 ) . Interestingly , one of 15 nef alleles ( clone 8 ) from another animal ( FBr ) was impaired in CD3 but not in CD4 or CD28 downmodulation ( Figure 9A ) . Since it differed in only eight amino acid residues from active FBr Nef alleles ( Figure S7C; Table S2 ) , we introduced all eight changes ( five individually and the RKT to HRV substitution in the “P-rich” region in combination ) into the FBr clone 8 Nef . Unexpectedly , only a homologous L123I change specifically enhanced its activity in CD3 downmodulation , while the reverse I123L substitution in the active FBr clone 6 Nef increased its ability to modulate CD3 even further ( Figure 9A; Table S2 ) . Thus , although the determinants of CD3 downmodulation by SIVsmm Nef are obviously complex and dependent on the specific Nef backbone , residue I123 seems to be specifically involved in Nef-mediated modulation of CD3 but not of other surface receptors . Further mutational analyses of SIVsmm Nef alleles differing in specific functions due to a limited number of amino acid substitutions should allow to pinpoint the critical residues .
In the present study we performed a detailed functional analysis of Nef proteins from naturally infected SM with different infection outcomes . Our results show that inefficient downmodulation of TCR-CD3 by SIVsmm Nef alleles correlated with the loss of CD4+ T cells . In other words , a more “HIV-1-like” Nef phenotype of SIVsmm , i . e . the inability to efficiently downregulate TCR-CD3 , was associated with a course of SIVsmm infection that was more reminiscent of pathogenic HIV-1 infection . At first view it may seem paradoxical that downmodulation of TCR-CD3 - a molecular complex essential for the function of T cells–may protect against the development of immunodeficiency . However , at least during chronic infection only a small fraction of CD4+ T cells is productively infected . Thus , it is conceivable that their functionality is not critical for the overall immune competence of the infected host . The rate at which these virally infected T cells die and must be replaced , however , must have a significant impact on the host's regenerative capacity . Furthermore , CD4+ helper T cells orchestrate the immune response and are typically able to exert significant effects on the immune system by relatively small numbers . Thus , virally infected activated CD4+ T cells likely contribute to high levels of immune activation by sequestering cytokines that induce the migration and inflammatory response of uninfected bystander cells . It is thus possible that a Nef-mediated suppression of CD4+ T cell activation and cell death contributes to a general reduction of infection-associated immune activation . Notably , our findings likely underestimate the real impact of Nef's ability to downmodulate TCR-CD3 on the rates of CD4+ T cell depletion because the functional differences between nef alleles from the CD4low and CD4high groups of SIVsmm-infected SMs are substantially more subtle than those between HIV-1 and the great majority of SIVs that are usually highly effective in removing TCR-CD3 from the cell surface [28] . In our previous study [28] , we observed a nonsignificant trend for more efficient MHC-I downmodulation by CD4high Nefs . The present analysis of a larger number of SIVsmm infections now reveals a significant correlation between efficient MHC-I downmodulation and stable high CD4+ T cell counts . A protective role of this Nef function is consistent with the observation that HIV-1 nef alleles from some long-term nonprogressors remain competent in downregulating MHC-I but are otherwise defective [61] , [62] . We also found that effective Nef-mediated MHC-I downmodulation is associated with low percentages of effector but high proportions of naïve and memory CD8+ T cells ( Figure 4G–I ) . This is in agreement with published data showing that weak MHC-antigen stimulation of CD8+ naïve precursors favors the generation of memory rather than effector CD8+ T-cells [63] and that SIVmac mutants selectively impaired in MHC-I downmodulation cause stronger CD8+ T-cell responses [26] . Finally , efficient downmodulation of both CD3 and MHC-I by Nef correlated with higher numbers of CD8+CD28+ T cells . While further studies are required to understand the functional significance of these associations , it is tempting to speculate that efficient MHC-I downregulation by Nef may contribute to the maintenance of high CD4+ T cells in naturally infected primates by two different mechanisms , i . e . by reducing CTL lysis of virally infected CD4+ T cells and by diminishing the levels of CD8+ bystander T cell activation . Accumulating evidence suggests that primate lentiviral Nef proteins generally promote viral persistance in the infected host by facilitating immune evasion ( i . e . by modulation of CD4 , MHC-I and Ii surface expression ) and by enhancing viral infectivity and replication [28] , [64] , [65] . Thus , Nef is most likely critical for the development and maintenance of high viral loads in natural as well as in recent and nonnatural SIV and HIV infections . However , Nef proteins from most SIVs and from HIV-2 also appear to exert a protective effect that is linked to their ability to downmodulate CD3 from the cell surface . As shown in the present study , lack of this Nef function is correlates with a decline in CD4+ T cells in primates that usually do not loose their CD4+ T cells as a consequence of SIV infection . Whether the reduced ability of Nef to downmodulate CD3 and MHC-I is the cause or consequence of CD4+ T cell decline remains to be investigated . However , there is no evidence that the quality or the potency of the antiviral immune response differs in SMs with high or low CD4+ T cell counts [37] , which argues against the possibility that a more effective immunity drives the observed functional differences in Nef . Our new data are thus in agreement with our previous hypothesis that the evolutionary loss of Nef-mediated downmodulation of TCR-CD3 contributes to the escalation of immune activation and the progression to AIDS in HIV-1-infected individuals [28] . The fact that SIVmac is pathogenic in macaques ( although its Nef downmodulates CD3 ) does not argue against this hypothesis . Firstly , this virus does not cause disease or high levels of immune activation when reintroduced into its original host , the sooty mangabey [6] , [66] , [67] . Thus , the pathogenic phenotype in macaques seems to be a consequence of an unusually high susceptibility of this nonnatural host species to SIV-induced disease rather than an inherent viral phenotype . Secondly , even in the SIV/macaque model increased T cell activation by Nef further accelerates disease progression [22] . Notably , nef-deleted attenuated HIV-1 infection is ultimately pathogenic in humans unless replication is completely suppressed [68]–[70] suggesting that Nef may accelerate disease progression in nonnatural HIV/SIV hosts mainly because it increases the viral loads by several orders of magnitude and not because it enhances virulence directly . These data suggest that in addition to maintaining high viral loads most primate lentiviral Nefs also reduce immune activation and T cell apoptosis and thus the damaging effects associated with these high viral loads by removing TCR-CD3 from the cell surface . We have previously shown that TCR-CD3 downmodulation by Nef inhibits induction of IL-2R expression , NF-AT activation and activation-induced cell death [28] . Here , we demonstrate that the efficiency of this Nef function also correlates with the suppression of the induction of PD-1 expression on virally infected T cells . The role of this receptor in the pathogenesis of AIDS has received a lot of attention because it has been shown that high levels of PD-1 expression correlate with declining CD4+ T cell counts and disease progression and are associated with reversible dysfunction of CD4+ and CD8+ T cells [52] , [53] . We thus anticipated that downmodulation of CD3-TCR would also inhibit the induction of PD-1 expression because this receptor is expressed predominantly on activated cells [54] , [55] . This was indeed observed ( Figure 6 ) . Furthermore , mutations in Nef disrupting CD3 downmodulation also impaired its ability to suppress PD-1 induction . Conversely , a truncated form of the SIVmac239 Nef ( tNef ) that is capable of downmodulating CD3 but inactive in modulating CD4 , MHC-I and CD28 and enhancing virus infectivity [50] , efficiently inhibits PD-1 induction ( data not shown ) . These data suggest that Nef-mediated downmodulation of CD3 may be required and sufficient for the suppression of PD-1 induction , although the possibility that these Nef mutants differ in additional functions cannot be excluded . It has been reported that PD-1 inhibits the proliferation of T cells and sensitizes them for apoptosis [46] , [71] , [72] . Thus , high levels of PD-1 expression should lead to increased programmed death of infected T cells and may hence accelerate their depletion . In agreement with this possibility we found that Nef alleles from SMs experiencing a significant loss of CD4+ T cells suppressed the induction of PD-1 less efficiently than those derived from animals with stable CD4+ T cell counts . Notably , upregulation of PD-1 on virally infected CD4+ T cells may also affect the function of bystander T and B cells via interactions with its ligands PD-L1 and PD-L2 [54]–[56] . Thus , further studies are required to clarify how differential levels of PD-1 expression on T cells infected with primate lentiviruses that do or do not downmodulate TCR-CD3 affect the immunological and clinical outcome of infection . In HIV-1-infected humans , high levels of proliferating CD4+Ki67+ T cells correlate with disease progression [73] , [74] . However , no correlation between the level of CD4+Ki67+ T cells and signs of disease progression or CD4+ T cell depletion was observed in SIV-infected SMs [37] . Unexpectedly , we found significantly higher numbers of proliferating CD4+Ki67+ T cells in animals infected with SIVsmm Nef variants that maintained high levels of CD4+ T cells despite the fact that Ki67 is usually seen as an activation marker . Since downmodulation of TCR-CD3 and MHC-I also correlates with reduced expression levels of CD95 on CD4+ T cells ( Figures 3D , S5B ) , the higher levels of CD4+Ki67+ T cells may simply reflect the prolonged survival of proliferating CD4+ T cells due to reduced levels of apoptosis . As discussed above , increased PD-1 expression is associated with reduced proliferative capacity [47] , [52] , [53] . Thus , it is tempting to speculate that Nef-mediated suppression of PD-1 may also contribute to the high levels of CD4+Ki67+ T cells . It has been proposed that enhanced levels of apoptosis of bystander CD8+ T cells are important for the development of AIDS in HIV-1-infected humans and SIV infected monkeys [46] , [75] . Thus , it will be of interest to investigate whether reduced CD95 expression on CD4+ T cells in SMs infected with SIVsmm variants that effectively downmodulate TCR-CD3 and MHC-I is associated with reduced apoptosis of bystander T cells expressing the CD95 ligand . A striking finding of the present study was that one naturally infected SM with very low CD4+ T cell numbers was the only one of 26 animals investigated which was infected with an SIVsmm strain whose Nef caused hyper-responsiveness of virally infected T cells to activation , a phenotype previously established for HIV-1 Nef alleles [31] . It is conceivable that direct enhancement of the responsiveness of virally infected T cell to activation by Nef may enhance activation-induced cell death and hence accelerate the loss of CD4+ T cells more dramatically than just inefficient inhibition of these processes . It is noteworthy , however , that this animal ( FYb ) did not show unusually high levels of immune activation and was already of old age [37] . Thus , further studies are required to clarify the role of Nef's ability to alter the responsiveness of infected T cells to activation in the course of primate lentiviral infection; i . e . it would be of high interest to assess whether infection of SMs with SIVsmm expressing the FYb nef allele experience a rapid loss of CD4+ T cells . Notably , the severe and persistent loss of CD4+ T cells in two of the four experimentally infected SMs examined was not only associated with inefficient downmodulation of TCR-CD3 but also with the emergence of SIVsmm variants capable to utilize CXCR4 and CCR8 in addition to CCR5 [38] . We have initiated studies to determine the temporal pattern of changes in Nef function and coreceptor usage and to clarify how the changes in these viral properties correlate with the loss of CD4+ T cells . Altogether , these data suggest that viral properties that are uncommon for SIVs but not for HIV-1 , i . e . an expanded coreceptor tropism and inefficient suppression of T cell activation , are associated with CD4+ T cell depletion even in naturally SIV-infected . Interestingly , none of the SIVsmm-infected SMs with persistent dramatic CD4+ T cell depletion developed AIDS [37] , [38] . Thus , further studies to elucidate which viral or host factors prevent disease progression in SIVsmm-infected SMs , even following dramatic CD4+ T cell depletion , are highly warranted . In summary , our data suggest that - in addition to other host and viral properties [1] , [2] , [76] - differences in specific Nef functions may modulate the immunological outcome of primate lentiviral infections . In particular , Nef-mediated downmodulation of TCR-CD3 and MHC-I seem to protect naturally infected SMs against loss of CD4+ T cells . Both of these functions presumably delay the apoptotic death of virally infected T cells , either by suppressing their activation or by preventing CTL lysis . However , these Nef functions may also affect the activation and programmed death of uninfected bystander T cells . The latter is supported by our new finding that efficient Nef-mediated MHC-I downmodulation is associated with low numbers of effector and high numbers of memory CD8+ T cells . At this point , our data are no definitive proof that there is a direct causal relationship between Nef-mediated downmodulation of CD3 and MHC-I and preserved CD4+ T cell homeostasis . Instead this will require the testing of selective SIV Nef mutants in vivo in appropriate SIV/monkey models , such as SIV-infected SMs or AGMs . Such experiments should also help to elucidate the role of T cell activation and/or CTL lysis in CD4+ T cell depletion and to assess whether preventing hyperactivation of the immune response would improve current treatment strategies .
Blood samples were obtained from 22 naturally and four experimentally SIVsmm-infected SMs , all housed at the Yerkes National Primate Research Center of Emory University and maintained in accordance with NIH guidelines . These studies were approved by the Emory University Institutional Animal Care and Use Committee ( IACUC ) . The virological and immunological characteristics of these animals have recently been described [37] , [38] . Plasma viral RNA was extracted using the QIAamp Viral RNA Kit ( Qiagen ) and nef genes were amplified by RT-PCR using primers SM-Nef-F1 ( 5′-GACAGATAGAATATATTCATTTCC-3′ ) and SM-Nef-R1 ( 5′-TCTGCCAGCCTCTCCGCAGAG-3′ ) . Sequence analysis of two clones per amplicon ( two per animal ) confirmed the integrity of the cloned nef alleles . Generation of HIV-1 ( NL4-3 based ) proviral constructs carrying functional nef genes followed by an internal ribosome entry site ( IRES ) and the eGFP gene has been described [25] , [28] . Derivatives containing stop codons at positions 73 and 74 of the nef ORF either alone ( nef* ) or in combination with mutations in the ATG initiation codon and two in frame stop codons at positions four and five of the nef ORF ( nef- ) disrupting the NL4-3 nef gene were generated by standard PCR and cloning techniques . Splice-overlap-extension PCR was used to replace the NL4-3 nef gene with the bulks of SIVsmm nef genes . Briefly , PCR fragments containing the 3′end of the NL4-3 env gene fused to the various pools of SIVsmm nef genes were cloned into pBR-NL43-IRES-eGFP-nef+ using the unique HpaI and MluI sites ( Figure S1 ) . Aliquots of transformed Escherichia coli were plated on Luria broth-ampicillin dishes to assess transformation efficiency , and the remaining 90% of the transformed bacteria were used for direct inoculation of medium-scale plasmid preparations . The percentage of the plasmid population containing an SIVsmm nef insert was estimated by restriction analysis and the integrity of all PCR-derived inserts was confirmed by sequence analysis . Nef alleles from 19 of the 26 SMs have been previously analyzed in transient transfection experiments [28] . Synthetic nef alleles , i . e . SMhi , SMlow , FCs clone 1mut and FBr clone 17mut , fused to the 3′end of the HIV-1 env gene were synthesized by Epoch Biolabs and cloned into pBR-NL43-IRES-eGFP vector using the unique HpaI and MluI sites . Jurkat and 293T-cells were cultured as described previously [28] . 293T-cells were maintained in Dulbecco's modified Eagle's medium containing 10% heat-inactivated fetal bovine serum . PBMC from healthy human donors were isolated using lymphocyte separation medium ( Biocoll Separating Solution , Biochrom ) , stimulated for 3 days with PHA ( 2 µg/ml ) and cultured in RPMI1640 medium with 10% FCS and 10 ng/ml IL-2 prior to infection . CD8+ T cells were depleted from the PBMC cultures using CD8 MicroBeads ( Miltenyi Biotec ) following the MACS separation protocol provided by the manufacturer . To generate viral stocks , 293T cells were co-transfected with the proviral HIV-1 constructs and a plasmid ( pHIT-G ) expressing the Vesicular Stomatitis Virus G protein ( VSG-G ) [25] , [28] . The latter was used to achieve comparable high initial infection levels for functional analysis . However , all HIV-1 constructs contained intact env genes and were thus replication competent following the first round of infection . The medium was changed after overnight incubation and virus was harvested 24 h later . Residual cells in the supernatants were pelleted and the supernatants were stored at −70°C . Virus stocks were quantified using a p24 antigen capture assay provided by the NIH AIDS Research and Reference Reagent Program . Virus infectivity was determined using TZM-bl and P4-CCR5 cells as described [65] . Briefly , the cells were sown out in 96-well-dishes in a volume of 100 µl and infected after overnight incubation with virus stocks containing 1 of p24 antigen produced by transiently transfected 293T cells . Two days post-infection viral infectivity was detected using the Gal screen kit from TROPIX as recommended by the manufacturer . β-galactosidase activities were quantified as relative light units per second ( RLU/s ) using the Orion Microplate Luminomter . CD4 , TCR-CD3 , MHC-I , CD28 and eGFP reporter expression in Jurkat T cells or human PBMC transduced with HIV-1 ( NL4-3 ) constructs coexpressing Nef and eGFP were measured as described [25] , [28] . CD8 , IL-2R , CD69 and PD-1 expression was measured by standard FACS staining , using CD8 ( BD Pharmingen , Clone RPA-T8 ) , CD25 ( BD Pharmingen , Clone M-A251 ) , CD69 ( BD Pharmingen , Clone FN50 ) and PD-1 ( BD Pharmingen , Clone MIH4 ) mAbs . For quantification of Nef-mediated modulation of specific surface molecules , the levels of receptor expression ( red fluorescence ) were determined for cells expressing a specific range of eGFP . The extent of downmodulation ( n-fold ) was calculated by dividing the MFI obtained for cells infected with the nef-minus NL4-3 control viruses by the corresponding values obtained for cells infected with viruses coexpressing Nef and eGFP . Human PBMC were first stimulated with PHA ( 1 µg/ml ) for 3 days . Subsequently , the cells were cultured in RPMI1640 ( 10% FCS , 10 ng/ml IL-2 ) , infected with various HIV-1 eGFP/Nef constructs and cultured for another 2 days . At this time the PBMC expressed very low levels of CD69 and IL-2R and hence had a resting phenotype . Thereafter , the PBMC were treated a second time with PHA and CD69 and IL-2R expression levels were measured by FACS analysis one and four days later . The frequency of virally infected apoptotic cells was determined using the AnnexinV ( AnV ) Apoptosis Detection Kit ( BD Bioscience ) as recommended by the manufacturer . Jurkat cells stably transfected with an NFAT-dependent reporter gene vector [31] were either left uninfected or transduced with HIV-1 Nef/eGFP constructs expressing various nef alleles . Except for those cells used as controls , cultures were treated with PHA ( 1 µg/ml; Murex ) . Luciferase activity was measured and n-fold induction determined by calculating the ratio between measured relative light units of treated samples over untreated samples as described previously [28] , [31] . The activities of nef alleles derived from SMs with low ( n = 15 ) or high ( n = 11 ) CD4+ T cell counts and of the subgroups of animals expressing nef alleles showing the highest ( n = 6 ) or lowest ( n = 6 ) activities in CD3 and MHC-I modulation were compared using a two-tailed Student's t test . The PRISM package version 4 . 0 ( Abacus Concepts , Berkeley , CA ) was used for all calculations . GenBank accession numbers for previously described SIVsmm nef sequences are DQ408682 to DQ408725 and for newly derived sequences EU636907 to EU636923 ( see Tables S1 and S2 for further detail ) . | The accessory Nef protein is commonly considered a “pathogenicity” factor of primate lentiviruses . However , SIVs do not cause disease in their natural hosts , although they all encode nef genes and sustain high levels of viremia . To better understand the role of Nef in natural nonpathogenic SIV infection , we compared the function of Nef alleles from two groups of SIVsmm-infected sooty mangabeys: ( i ) those that maintained normal CD4+ T cell counts and ( ii ) a small subset ( 10%–15% ) of animals that exhibited a considerable loss of CD4+ helper T cells . We found that the efficiency of two specific Nef functions , i . e . , downmodulation of TCR-CD3 and MHC-I , correlated with preserved CD4+ T cell homeostasis , as well as with other immunological features , such as high numbers of proliferating CD4+ Ki67+ T cells . Moreover , lack of CD3 surface expression was associated with low levels of apoptosis and PD-1 expression by virally infected T cells . Thus , the ability of Nef to remove TCR-CD3 and MHC-I from the cell surface may help the natural hosts of SIV to maintain normal CD4+ T cell counts despite high levels of viral replication by preventing activation-induced cell death and CTL lysis of infected T cells and/or CD8+ T cell activation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
]
| [
"virology/virulence",
"factors",
"and",
"mechanisms",
"virology/virus",
"evolution",
"and",
"symbiosis",
"virology/immunodeficiency",
"viruses",
"virology/mechanisms",
"of",
"resistance",
"and",
"susceptibility,",
"including",
"host",
"genetics",
"virology/immune",
"evasion"
]
| 2008 | Inefficient Nef-Mediated Downmodulation of CD3 and MHC-I Correlates with Loss of CD4+ T Cells in Natural SIV Infection |
Neural processing rests on the intracellular transformation of information as synaptic inputs are translated into action potentials . This transformation is governed by the spike threshold , which depends on the history of the membrane potential on many temporal scales . While the adaptation of the threshold after spiking activity has been addressed before both theoretically and experimentally , it has only recently been demonstrated that the subthreshold membrane state also influences the effective spike threshold . The consequences for neural computation are not well understood yet . We address this question here using neural simulations and whole cell intracellular recordings in combination with information theoretic analysis . We show that an adaptive spike threshold leads to better stimulus discrimination for tight input correlations than would be achieved otherwise , independent from whether the stimulus is encoded in the rate or pattern of action potentials . The time scales of input selectivity are jointly governed by membrane and threshold dynamics . Encoding information using adaptive thresholds further ensures robust information transmission across cortical states i . e . decoding from different states is less state dependent in the adaptive threshold case , if the decoding is performed in reference to the timing of the population response . Results from in vitro neural recordings were consistent with simulations from adaptive threshold neurons . In summary , the adaptive spike threshold reduces information loss during intracellular information transfer , improves stimulus discriminability and ensures robust decoding across membrane states in a regime of highly correlated inputs , similar to those seen in sensory nuclei during the encoding of sensory information .
The essential computation performed by neurons is the ( non-linear ) integration of synaptic inputs and subsequent generation of a spike . This summation is determined by multiple factors , including the passive membrane properties , active currents , and the neuronal geometry [1–3] . The currents at the axon hillock have voltage-dependent activation dynamics , which accelerate and then self-sustain the depolarization , eventually leading to a spike . The voltage at which this acceleration occurs is commonly referred to as the voltage threshold of spiking or spike threshold . Recently , several studies have empirically demonstrated that this threshold is not constant as recorded in cortical [4–8] and in subcortical [2 , 9 , 10] neurons . Rather , the spike threshold exhibits a dependence on the slope of the depolarization , such that fast depolarizations produce spikes at lower thresholds , and slower depolarizations at higher thresholds . The relationship is monotonically decreasing over a range of a few millivolts . This phenomenon is qualitatively different from the thoroughly studied spike-frequency adaptation , which acts on longer time scales [11–13] and depends on suprathreshold activity . The effects of the adaptation investigated here , will manifest already before the ‘first’ spike , e . g . at response onset . Recently , Fontaine and colleagues [2] have demonstrated that this threshold dependence can be captured using a threshold which itself ‘chases’ the membrane potential , however , does not instantly adapt to the new membrane potential but approaches its new value with a certain time-constant . Consequently , if one considers two fluctuations of equal size , but different speed , the faster fluctuation will encounter a lower threshold ( still in the approach of its final value ) and thus already lead to a spike , whereas the slower fluctuation may encounter a higher threshold and thus lead to a spike at a greater depolarization ( or fail to spike ) . Hence , the propensity to spike depends on the recent history of the membrane potential and in particular on the most recent rate of depolarization leading up to a potential spike . Multiple mechanisms could underlie this modulation in spike threshold , including adaptive changes in sodium channel inactivation [5–7 , 14] and potassium conductances [4 , 15] . Given that the existence of these conductances is common among spiking neurons , an adaptive spike threshold is likely to be ubiquitous features of neurons . An adaptive threshold could have important computational consequences , since this would rapidly modulate the steepness of the input-output function of neurons [14] . Several authors [2 , 5–7 , 15] have speculated on the possible relevance of an adaptive threshold in the sense described above , however , to our knowledge , no study has thoroughly investigated its computational consequences . Hence , here we set out to evaluate the effect of an adaptive threshold on the level of single neurons in feedforward excitatory networks in the context of stimulus processing . We investigate how a single neuron’s response to different rates and patterns of inputs varies as the action potential generation is gated by fixed , i . e . constant , or adaptive thresholds . Furthermore , we investigate how the state of the neuron ( resting membrane potential prior to stimulus onset ) influences the response behavior using both recordings from barrel cortex as well as neuronal simulations . The results suggest that neurons with adaptive threshold perform better than with a constant threshold when the stimulus is encoded by highly correlated inputs . Such a neuron encodes the stimulus more robustly , even in the context of noisy fluctuations of the membrane potential . For the state differences , we find the temporal reference used for decoding—internal or external—to play a major role: If the time reference is computed internally based on the timing information from the correlated spiking of local populations , an adaptive threshold turns out to be beneficial . Together these results show the usefulness of adaptive thresholds in performing temporally precise and robust computations .
While the adaptiveness in threshold is typically described in relation to the slope of the membrane potential [5–7 , 9] ( Fig 1 ) , for computational purposes it is more insightful to describe it in relation to the properties of the synaptic input to a single compartment postsynaptic neuron . On a mechanistic level , Fontaine et al [2] identified the membrane potential itself as a valuable predictor of the threshold , which then also defines the adaptation of the threshold in the neural model ( see Materials and Methods ) . Subthreshold dynamics are expected to differ slightly between adaptive and fixed threshold , as the threshold dynamics ( Methods , Eq 1 ) are already active below threshold . While this choice is practically useful , it does not—or only indirectly—allow one to address questions such as the dependence of the spike threshold on the input rate , pattern , or correlation . Therefore , we first substitute the amplitude and slope of the incoming EPSP , by the number of synaptic/neuronal inputs Ninputs and their temporal dispersion σ , which parametrizes the Gaussian distribution from which the EPSP arrival times are drawn . The parameterization of the input by Ninputs and σ is valid , since these two parameters predict the amplitude ( Fig 2A1 and 2B1 ) and slope ( Fig 2A2 and 2B2 ) of the resulting EPSPs with little variance ( dark surface in the bottom of each panel ) . The quality of the prediction and the shape of the dependence are largely independent of whether the adaptive or the fixed threshold model is considered ( Fig 2A1 vs . 2B1 and 2A2 vs . 2B2 ) . Subthreshold dynamics are expected to differ slightly between adaptive and fixed threshold , as the threshold dynamics already act on the membrane equation ( Eq 1 ) below threshold . This is in contrast to threshold models in which the spike threshold is not part of the membrane equation ( e . g . leaky integrate and fire neuron ) , where the adaptation of spike threshold does not influence the subthreshold dynamics . As expected , increases in Ninput and decreases in σ increase both the amplitude ( AEPSP ) and the slope ( SlEPSP ) of the EPSP . Hence , the values of Ninputs and σ map approximately bijectively ( one-to-one ) to corresponding values of AEPSP and SlEPSP and thus span the space of EPSPs w . r . t . to these two experimentally relevant parameters . Based on these parameters , we investigated the transition between sub- and suprathreshold activity via the spike probability . Throughout this study the only source of variability in the simulations is trial-to-trial variation in synaptic strength ( including failures , see Methods ) , with the exception of an investigation of noise susceptibility ( see below ) , where Poisson spikes are added to the presynaptic input . The adaptive threshold model exhibits overall a steeper transition to spiking as a joint function of σ and Ninputs than the fixed threshold model ( Fig 2C1 vs 2C2 ) . Analyzing the two parameters separately , however , shows that the transition is only steeper for the temporal precision of the input , σ ( Fig 2D1 vs 2D2 ) , but not for the input rate , Ninputs ( Fig 2E1 vs 2E2 ) . The dependence of spike probability on σ and Ninputs exhibited similar shapes: Spike probability as a function of only σ , is well fitted by a sigmoid , with the adaptive model exhibiting a steeper slope as a function of different σ’s ( adaptive: s = 0 . 16 , fixed: s = 0 . 28 ) . Fitting sigmoids to the spike probability as a function of the Ninputs shows an inverse relationship in steepness between the adaptive ( s = 0 . 98 ) and the fixed ( s = 0 . 63 , E1 vs . E2 ) threshold neuron . In addition , the midpoint of the transition is located at a lower σ for the adaptive than for the fixed threshold ( 2 . 6 vs . 3 . 5 ms ) , while the midpoints for the dependence on Ninputs differ only slightly ( 34 vs . 37 inputs ) . The latter is a consequence of setting the fixed threshold matched to the average of the adaptive threshold . The adaptive threshold model is thus able to respond to a wider dynamic range of the number of inputs contributing to a stimulus , restricted to a smaller range of integration time-windows ( given here by the lower lowpass limit w . r . t . σ , compare Fig 2D1 to 2D2 ) . Neural integration projects a high-dimensional input onto a lesser dimensional membrane potential and eventually onto a binary time series as spikes . While a certain amount of information will always be ignored in this process , the goal of mutual information analysis is to quantify the ability of neurons to give distinguishable responses to different inputs . In the following , we will focus on two encoding schemes on the input side , population rate and population pattern ( which will be abbreviated as rate and pattern below ) . In the case of population rate encoding ( Fig 3 top ) , the studied neuron receives input from a population of presynaptic neurons , and inputs differ in their overall firing rate . Although each presynaptic neuron generates at most one spike in a given trial , the input rate is varied by changing the number of neurons that contribute to the current input . The spikes from the presynaptic neuron population are normally distributed as a function of time around a common average time t0 with standard deviation σ . Spike times are drawn randomly for each trial , while only the number of participating neurons stays the same across trials for a given input . Under this encoding schema , the only difference among stimuli is the population firing rate of the input neurons , and the temporal patterns of presynaptic spikes provide no information . For the simulations reported , in total 11 different stimuli were given ( range of active neuron number , 40–60 , incremented by 2 ) , thus the total stimulus entropy was log2 ( 11 ) ≈3 . 46 bit . Adaptive threshold neurons performed slightly better for low σ , but substantially worse for larger σ for rate encoded inputs ( Fig 3A ) . In these simulations noise exists as variability in the synaptic weights ( modeled as described in Feldmeyer et al . 2002 ) , but no additional inputs were added . Adding additional independent Poisson noise inputs generally reduces the recoverable mutual information ( Fig 3B1 and 3B2 ) , but does not change the qualitative finding of a better decoding of rate by fixed threshold neurons for larger σ ( Fig 3B3 ) . Overall , both adaptive and fixed threshold neurons exhibit a low pass-behavior in σ , which results from the fact that a greater σ at some point exhausts the integration window of each model . In the case of the population pattern encoding ( Fig 3 bottom ) , the different inputs are created by setting distinct predetermined patterns across neurons , which all have a fixed population firing rate ( Ninput = 50 ) . Under this condition , the same number , but a different subset of neurons is activated by different stimuli . Across trials , each presynaptic neuron gives the same spike time for a given stimulus . Hence , distinction between different stimuli has to rely on the temporal differences between different input spike patterns . As above , 11 different stimuli were delivered , resulting in ~3 . 46 bit stimulus entropy . Adaptive threshold neurons perform substantially better for low and worse for greater σ . Adding independent Poisson noise changes the relation between the models qualitatively , with the clear difference between them progressively giving way to a weak , but inverse relationship ( Fig 3D ) . This is due to the dependence of the MI peak on the noise strength: in the adaptive threshold model larger noise shifts the peak more quickly to greater σ's than in the case of the fixed model ( Fig 3D1 vs . 3D2 ) . Both model neurons exhibit a band-pass behavior w . r . t . to σ , in contrast to the low-pass behavior for rate encoding , which follows from the fact that patterns with very small σ effectively get squeezed to the same , single-time bin pattern . In summary , adaptive threshold neurons are useful encoders for precisely timed , or well correlated , inputs . Interpreted inversely , the shortened integration window of the adaptive threshold model , allows it to ignore inputs outside a certain period , and thus make decisions more rapidly ( see also [16] ) . On the other hand , a constant threshold would be advantageous if the processing has to happen over a wider range of temporal correlations in the input . We have performed a limited set of simulations of Hodgkin-Huxley neurons with a range of parameters , which transitions its behavior from an adaptive to a fixed threshold ( similar to [17] ) . Overall , the Hodgkin-Huxley model performed quite similar to the simpler model described above ( S1 Fig ) . The adaptive threshold is governed by a few parameters ( see Eqs 2 and 3 in Methods ) , which determine its dynamics and asymptotic value . Next , we address the interplay of the temporal dynamics of the threshold , given by the time constant τθ ( Eq 2 ) , with the passive integration of the neuron , given by the membrane time constant τm ( Eq 1 ) . For this purpose , we repeated the simulations above for a matrix of τθ and τm values , and evaluated the effect on firing rate , represented information and most informative temporal dispersion σ . For rate encoded stimuli , the average firing rate was limited similarly by both τθ and τm ( Fig 4A and 4B , in plots A-D and F-I the other time constant is set to 4 ms ) . The reason differs slightly between them: while small τm prevent integration for larger σ , small τθ allow the threshold to change rapidly , adapting fully to the input and thus preventing the generation of a spike . The represented information was limited by the firing rate and thus affected similarly by both time constants ( Fig 4C and 4D ) . Overall , MI therefore follows a lowpass behavior w . r . t . σ , although certain combinations of time constants with σ led to transient bandpass characteristics ( e . g . for τθ = 2 ms , where the bin size of the MI analysis cannot distinguish the resultant spike times any more ) . The joint influence of τθ and τm on temporal integration exhibited a monotonic increase of σcm ( Fig 4E ) , which was defined as the center of mass in MI w . r . t . σ for each choice of τθ and τm , computed as the average over σ , weighted by the normalized MI . For pattern encoded stimuli , the dependence of firing rate was analogous to the rate encoding case , exhibiting a low-pass limited by both time constants ( Fig 4F and 4G ) . The σ of peak MI increased in correlation with the time constants , broadening alongside ( Fig 4H and 4I ) . The joint dependence of σcm on τθ and τm exhibited a similar dually monotonic increase ( Fig 4J ) , as predictable from the individual dependencies . Overall , the best time scale for integrating input varies smoothly and similarly with the governing time constants τθ and τm . Further , this relation is independent of the input encoding scheme . We also investigated the influence of the input population's size ( see S2 Fig ) . Large input populations were favorable for rate encoding; small ones for pattern encoding , however , adaptiveness of the threshold did not have an effect on the Ninput dependence . Cortical populations undergo membrane state changes , depending on the wakefulness or attentional context of the brain [18 , 19] . The state changes manifest themselves on the single cell level as a shift in the subthreshold membrane potential , thus bringing each cell to a different voltage distance to its spiking threshold . While the sources of the state changes are not fully understood , their presence will either change network computation , or require network computation to be robust against the changes . An adaptive threshold could be a contributor to robust processing under different states . We therefore investigated the robustness of adaptive compared with fixed threshold neurons for different initial membrane voltages for rate-encoded inputs ( Ninput range 50–60 , with increment step size of 2; in total 6 stimuli with log2 ( 6 ) ≈ 2 . 58 bit entropy ) . An ideal observer could decode the stimulus from the spiking activity of the simulated neuron , with or without knowing the resting membrane potential state of the simulated neuron . We defined a ( relative ) robustness index RI ( see Methods ) to quantify the neuron’s robustness to state changes , which was defined as the ratio between I ( S;R ) and I ( S; R , state ) . I ( S;R ) is the stimulus information , which can be gathered from the neural response without knowing the state , i . e . the information robustly present in the neural response independent of state knowledge , and I ( S;R , state ) , the stimulus information that can be gathered when also knowing the state . A neuron that encodes perfectly robust to state-differences would not show any improvement by knowing the state , and thus have an RI close to 1 . The values of RI typically fall in [0 , 1] ( in some cases the maximal value is exceeded due to numerical imperfections ) , since adding state-information increases the mutual information due to part of the variability being explained by state knowledge ( pointed out for example by [20] ) . The adaptive threshold model exhibited a reduced dependence of response latency as a function of different initial states ( Fig 5A , the shown spikes are relative to presynaptic stimulus onset ) . For a small difference of only 3mV ( -65 vs . -62mV , Fig 5A1 ) the spike times of the adaptive threshold neuron are almost identical across states ( red , different shades indicate different Ninput , with more solid colors corresponding to greater Ninput ) , while the times are already noticeably shifted for the fixed threshold neuron ( blue ) . Correspondingly the mutual information with state knowledge ( Fig 5B1 , solid colors ) , is only slightly larger than the mutual information without state knowledge ( light colors ) in the adaptive case , but substantially larger in the fixed case . The adaptive model has an RI close to 1 , while the fixed model’s RI is much lower , which remains fairly constant as a function of σ ( Fig 5B1 bottom ) . If one considers larger differences in state ( Fig 5A2 , -72 vs . -62 mV ) , the changes in spike time due to the state changes become more severe . Especially for the fixed threshold model , the set of spike times becomes essentially disjoint , and in the more hyperpolarized state ( -72 mV ) the dispersion of the spike times increases substantially . In addition , failures of spike elicitation are observed ( Data on top , marked as NS ) . For the adaptive threshold model the spike time shift becomes stronger than observed with smaller ( 3mV ) state difference , but stays limited in comparison to the fixed model . Interestingly , from an information theoretic perspective , this leads to an inversion of the situation: The fixed model is now less dependent on the knowledge of state ( Fig 5B2 , top ) , again almost independent from σ ( Fig 5B2 , bottom ) . The basis of this inversion is that decoding across states can also work optimally , if the response time distributions are different , but disjoint across states . The difference in robustness against variations in initial membrane potential of the models observed holds more generally across a wider range of state differences . The differential behavior described above rests on the overlap between the response distributions across different states ( Fig 5C , quantified by the correlation coefficient across PSTH-bins CCPSTH ) . For small state differences the PSTHs are almost identical , leading to a correlation coefficient close to 1 . For large state differences the PSTHs are disjoint and hence the correlation coefficients are close to 0 . In between , the fixed threshold model exhibits a faster decay in correlation coefficient , thus indicating a faster drift of onset times . Qualitatively , the correlation coefficient between PSTHs is a good predictor for the resulting robustness index ( RI ) during decoding ( Fig 5D ) . High RI , i . e . high robustness , is achieved both for highly similar ( CCPSTH = 1 ) and dissimilar responses ( CCPSTH = 0 ) . Correspondingly , RI varies non-monotonically as a function of state difference , with the adaptive threshold neuron being more robust for small state differences , and the fixed threshold neuron being more robust for larger state differences ( Fig 5E ) . Hence , an adaptive threshold improves the robustness to state differences in comparison to a fixed threshold only for a limited range of small state differences . For larger state differences , the confusion generated by small shifts in timing is outweighed by the decoding possible across different time-windows . This occurs in the fixed threshold case , where response times shift far enough across states to make them nearly disjoint , thus in principle providing it with an advantage in information transmission . Classically , information theoretic decoding is performed in reference to the timing of the stimulus , e . g . spike-times are computed in relation to the time of stimulus onset ( ‘stimulus-centered’ , [21–23] ) . While this provides an objective reference , the brain's internal decoding mechanisms may not have direct access to the ( external ) information about stimulus onset . Alternatively , responses can be decoded in relation to the response timing of other neurons [24 , 25] . Use of this 'response-centered' decoding is not only closer to the brain's internal perspective , but also provides more accurate decoding if the time of stimulus onset is uncertain [25] . In the case of response-centered timing , the adaptive threshold model is found to be generally more robust to differences in state than the fixed threshold model . The internal reference time was defined as the peak-time of the population response ( termed 'columnar synchronous response' in [25] ) . Since the neural population encompasses a range of tuning preferences , this leads to an average response time of the neural population to the stimulus set ( see Materials and Methods ) . Effectively , the spike-times of the analyzed neuron are thus shifted to a common average response time of all simulated neurons , which will , however , depend on the membrane state and neural response properties , e . g . threshold type ( see Fig 6A ) . For small differences in state , both adaptive and fixed threshold models show little difference in their response time and variation in response time ( Fig 6A1 , red vs . blue , different shades indicate different stimuli , as in Fig 5 only rate encoding is considered here ) . Consequently , the RI is quite high ( Fig 6B1 , especially bottom ) . For larger state differences , the response time distributions differ strongly in their variance , with the fixed threshold model exhibiting a larger increase in spread for the more hyperpolarized state ( Fig 6A2 ) . The reduced variance across states and stimuli here exemplifies the effect of the adaptive threshold . The quality of unified decoding across states deteriorates for the fixed threshold model , thus leading to a larger information gain from the inclusion of state information , as indicated by the reduced RI values ( Fig 6B2 ) . As a function of state-difference , both the fixed and the adaptive threshold model exhibited better conserved PSTHs across states ( compare Fig 6C with Fig 5C ) when ‘response-centered’ reference was used . However , the adaptive threshold model profits more and thus a wider gap between fixed and adaptive threshold model results in terms of PSTH correlation coefficients . As before , the correlation coefficient between the PSTHs remains a good predictor for the RI values ( Fig 6D ) , although the shape slightly differs from the previous case ( Fig 5D ) . For 'response-centered' decoding , the robustness across states is generally higher for the adaptive than the fixed threshold model ( Fig 6E ) , independent of the size of the state-difference ( compare to Fig 5E ) . While the robustness partially depends on the particular combination of states ( combinations averaged in Fig 6E ) , the variability is small compared to the effect size ( errorbars in Fig 6E ) . In summary , an adaptive threshold neuron can be decoded more robustly than a fixed threshold neuron , if the stimulus timing is known from the brain's internal perspective . Under these circumstances , the absolute timing is converted to a relative timing in the population , and the reduction in response variance in the adaptive threshold model renders responses across different states more comparable . Cortical neurons have been shown to display spike threshold adaptation [4–8] . However , the effect of this adaptive threshold on the information transmission across resting membrane potential states has not been quantitatively investigated . We collected whole-cell patch clamp recordings from pyramidal neurons in L2/3 mouse barrel cortex in acute slice preparations , and delivered the stimuli after clamping the somatic membrane potential across different resting potentials . The stimuli consisted of currents with different rising slopes injected from a bipolar electrode placed in the L4 of the same barrel column as recorded cells , which were comparable to the “rate encoding” in the input population as mentioned earlier . In total 4 different stimuli were delivered , with ~2 bits of total entropy . Similar to previous studies , the recorded neurons exhibited an adaptive threshold that depended on the slope of the input for all states ( -80 , -70 , -60mV; r2 = 0 . 60±0 . 18 , 0 . 45±0 . 13 , 0 . 52±0 . 10 , respectively . Values are mean±s . d . , n = 11 ) investigated ( Fig 7A and 7B ) . Interestingly , this relationship appears not to be strongly dependent on the resting membrane potential , as the average slopes obtained from linear regression across the different states were comparable , i . e . -0 . 99 ( 0 . 27 ) , -0 . 92 ( 0 . 19 ) , and 0 . 90 ( 0 . 23 ) ms respectively ( p = 0 . 48 , one-way ANOVA with correlated samples , numbers in parentheses are s . d . ) . The neurons also exhibited a high PSTH similarity , as well as a high robustness of information encoding across states , reminiscent of the adaptive threshold model . For small state differences , both the response patterns ( Fig 7C1 ) , the decoded information and the RI value remain comparable ( Fig 7D1 ) across stimulus-centered ( orange ) and response-centered ( red ) decoding . For larger state differences the improved robustness of decoding to state variation with a response-centered threshold becomes evident ( Fig 7D2 ) . The correlation coefficient between PSTHs at different states remained high for a wide range of state differences . This range was even wider than the one for the adaptive model for response-centered decoding ( compare Fig 7E to Fig 6C ) . This difference between real and simulated neurons could either stem from a mismatch of the model parameters to the real neuron or an additional mechanism not accounted for by the model . Overall , the relation between across-state robustness ( as indicated by the RI ) and correlation coefficient ( Fig 7F ) was similar for real and model data ( compare orange to Fig 5D , and red to Fig 6D ) . Robustness across states exhibits a shape that stays closer to the adaptive than the fixed threshold model behavior . It rises slowly and decays much later , especially for the static decoding ( Fig 7G compared to Fig 5E ) . In summary , the information transmission behavior of cortical neurons exhibited similar features of robustness across states as the adaptive threshold model . An adaptive threshold could therefore be a mechanism which contributes to achieving this robustness , providing a functional reason for its widespread existence in the mammalian nervous system .
Threshold adaptation has been observed in a range of systems and modalities , and therefore appears to be the norm rather than the exception . In cortical neurons , the spike threshold is shown to be correlated with the average membrane potential prior to a spike [6 , 26] , as well as inversely correlated with the rate of membrane potential depolarization in both excitatory [5–8] and fast-spiking interneurons [4] . The adaptive threshold increases the sensitivity to synchronized presynaptic inputs , while suppressing uncorrelated inputs , thus potentially increasing stimulus selectivity . Similar forms of spike threshold adaptation have been reported in the thalamus [9] , the subthalamic nucleus [10] and the auditory brainstem [2] . Importantly , threshold adaptation has to be separated from spike frequency adaptation , which occurs as a consequence of spiking , but is not influenced by the subthreshold voltage . Further , the present study skips the synaptic current level , but works directly on the neural output , which is generated by the nonlinear spiking process . Previous studies have speculated and partially linked the adaptive nature of the spike threshold to improvements in neural coding , but not performed corresponding analyses to quantitatively test this hypothesis . The clearest investigation of the mechanism is found in Fontaine et al . [2] , where the difference between the adapting threshold and the membrane potential is identified as the 'effective signal' , which determines under what conditions the neuron will spike . They demonstrate directly , in the context of neurons from the barn owl's inferior colliculus , that the response selectivity shifts towards temporally more tightly tuned coincidences . The present reparametrization of the EPSPs in terms of the input temporal precision is partially guided by this finding . Their analysis , however , does not proceed to a full decoding analysis as presented here . We recover their result in a more general form for the information analysis ( Fig 3 ) , with improved encoding for tightly tuned inputs in the adaptive threshold neurons than in the fixed threshold neurons , both with respect to rate and pattern encodings in the input . Another interesting perspective is provided by the study of [Fontaine et al 16] , which investigated the transition from encoding sound pressure to encoding the envelope of sound in the barn owl's cochlear nucleus . Remarkably , they find that this is enabled by the adaptive threshold of neurons in the nucleus angularis . Hence , the adaptive threshold can also take the role of transitioning between codes , rather than changing the precision within a code . While the functional significance of state changes is still debated , research over the last decade has indicated that ( subthreshold voltage ) state changes occur under a variety of conditions , ranging from different states of wakefulness [18] , movement [27] or task involvement [19] , to rapid changes during sensing [28] . While the activity of neurons will undoubtedly be modulated by the change in subthreshold membrane potential , it is not clear how the decoding will be affected . We presently addressed this question on the basis of modeled and real data , by computing the mutual information between stimulus and response for responses from neurons in different states . In particular , we investigated how insensitive the decoding was to the subthreshold state , by comparing decoding with and without state knowledge . The results suggest that an adaptive threshold confers an increased robustness across states , especially if the decoding is performed based on an internal time reference . As we argue below , a decoding scheme consistent with the internal perspective of the nervous system has to rely solely on quantities which are internally available . Consequently , an adaptive threshold provides support to the idea that the processes of decoding can remain unchanged , even if the state changes . While other mechanisms ( e . g . the hyperpolarization activated cation currents [29] ) may also contribute to robust processing , an adaptive threshold may partially explain the relative constancy of perception across different states . In a related study , Safaai et al [20] have investigated how much information about the stimulus can be gained by being aware of the current state . Their measure accounted for the state-dependent variability within the total stimulus information . The present robustness index RI constitutes 'the other side of the coin' , in asking how consistently a cell responds across states , i . e . how much information is lost in not knowing the state . Neural communication occurs in time , and it is not clear if there is a common clock in the brain which could be used to determine the start and end of a message . The interpretation of a message depends on the temporal reference used , and can determine the meaning of the message , as well as the set of messages distinguishable , and hence the capacity of the channel . For example , a given spike pattern "1011" ( binned over time ) can be interpreted as "0101100" and "0001011" depending on the time reference . These messages may carry different information . This becomes particularly evident in the typical scenario of neural integration of messages ( spike-trains ) from multiple sources at the same time , where a time-shift between two spike-trains will affect their integration by the post synaptic neurons [30 , 31] . Since a neuron only has access to information from its own inputs ( including modulatory inputs ) , a time reference has to be generated from the brain's internally available information . A time reference can be generated in multiple ways . For example , in vocal communication , pauses between words or sentences are used as markers to define starting points for interpreting parts of the entire message . Given the onset response properties of sensory neurons , these pauses will generate volleys of spikes , which can be used as a temporal reference . As suggested by Panzeri and Diamond [25] , a similar interpretation could hold for the onset responses generated by tactile events in the somatosensory system [32 , 33] as cortical neurons integrate spatiotemporal information on behaviorally relevant time scales [34] . More generally , a given neuron could derive a time reference from the average timing of its synaptic input . The existence of such timing references is more generally suggested by the existence of large-scale oscillatory signals both on the cortical [35–37] and subcortical level . As demonstrated here , an adaptive threshold provides advantages in information representation , since the variance of the response time of a neuron with an adaptive threshold is more restricted than that of a neuron with a fixed threshold , both across stimuli and across subthreshold initial states . In combination with the use of a population-based time reference , this attributes a role in generating robust information transmission to the use of an adaptive threshold . Conversely , due to its adaptation to state , the adaptive threshold model may have less access to the state itself , which is e . g . valuable if the state value should play a role in the processing of information . This trade-off between fixed and adaptive encoding has been observed before in multiple contexts ( e . g . [38 , 39] ) . Information theoretic methods are useful to evaluate and compare different decoding mechanisms , since they are quite general , make only few assumptions and lead to objective performance estimates . Results from a decoding method are , however , only relevant , if they are compatible with the processing performed in subsequent stations in the brain . Below we address some some limitations/caveats stemming from information theory in general and relating to the present study . First , we used the entire spike train ( binned at several temporal precisions ranging from 0 . 5–30 ms ) as the response of the neuron . While this approach guarantees that we used all the available information , it also assumes that the decoder can wait for the entire time-span to decide whether and how to respond . A closer approximation to the decoding performance of a ( point ) neuron would be a decoder with a limited integration-time and a memory term , that weights recent inputs stronger than past inputs . As the processing is performed online and in real-time , multiple decoders of this kind could be lined up in sequence to perform classification of longer responses . Alternatively , one could assume that ( internal ) decoding only happens on short timescales and thus restricts the view to a single decoder . This would result in information loss , but it will not significantly affect the computational power of the adaptive threshold over the fixed threshold as described herein . Second , the present comparison between adaptive and fixed threshold neurons with respect to decoding across states assumed that the responses from all states were available to the decoder . While this assumption is reasonable to assume over long time-scales , one could postulate that a decoder is instead only matched to one state ( e . g . the up state ) , in which case robustness of the decoding strategy could then be defined as how well the state-specific representation can transfer to other states . Concretely , the decoder would specifically decode ( i . e . be trained ) for one state , and loss in decoding quality in another state would depend mostly on the similarity of the response between states . In this case we would predict a neuron with adaptive threshold to still encode more consistently across states , since a state change would lead to less changes in neural response . Generally , the robustness in decoding across states could be achieved by either an adaptive threshold mechanism which leads to less changes in neural responses to state variations ( as shown in current study ) , or by using the trial-by-trial state knowledge to discount part of the neural response variability caused by state changes ( see [20] ) . Safaari et al . [20] modeled the effect of state variation on neural response explicitly , thus their approach gives a lower bound on the information gain by knowing the states; here by using mutual information analysis we gave the upper bound of the information gain the state knowledge could contribute . The integration and transfer characteristics of individual neurons are relevant as they form the basis of information processing on the network level . The adaptive threshold neuron's sensitivity to correlated inputs could contribute to a much discussed property of neural representation , namely sparseness . Limiting the neuron’s response to correlated input reduces its responsiveness , so it responds only to a smaller subset of input patterns/rates , thereby reducing the overall number of spikes in the network . Hence , a side-effect could be a more energy efficient operation based on the limited number of spikes . The adaptive threshold neuron's general ability to follow changes in preceding voltage reduces the variability in the spike latency after stimulus onset as compared with a neuron with a fixed threshold . Consequently , when the same stimulus is presented , the response variability with respect to state differences is reduced ( see Fig 5A ) . Hence , in a network with multiple layers , the dispersion of spike timing would accumulate at a lower rate ( per neuron ) . To make use of this precision , as discussed above , the time-frame for decoding should not be fixed to the presynaptic response onset , but rather be defined by the average timing on the postsynaptic side [25] . While certain consequences of single neuron properties to the network level can be predicted well , explicit simulations or theoretical explorations are required to test these hypotheses in further detail . In the present study , an adaptive spike threshold was modeled on the basis of a phenomenological description by Fontaine et al . [2] , which was shown to make accurate spike-time predictions for the case of subcortical neurons in the inferior colliculus of the Barn owl . Alternative descriptions of an adaptive threshold could be based on biophysical mechanisms , such as the inactivation of Na+ channels [14] and the activation of K+ channels ( e . g . low-voltage activated Kv1 channels [15] ) . These descriptions would be directly based on a biological mechanism and could therefore be tested more directly using specific tools , for instance . Na+-channel blockers together with a dynamic clamp-based electrical insertion of voltage-dependent conductances [40] . However , for the present purpose , i . e . a neural decoding analysis , the phenomenological properties of the model are more relevant than its biophysical basis . As long as the Fontaine model is consistent on the phenomenological level , the present results should transfer accurately to more biologically realistic models . We used the present model because of the usual advantages of using simplified models: simplified simulation and a more direct link between an underlying mechanism and its computational consequence . In the opposite direction , despite the elegance of the Fontaine model , one may ask whether computationally simpler models could account for the adaptive threshold , which would be advantageous for large scale simulations . While the present work addresses many questions relevant to the computational effects of an adaptive threshold , a range of questions remains open . In particular , including other neural constraints ( e . g . limited memory ) would be of interest , as well as the explicit study of network activity , and the consequences on sparseness and efficiency on the network level . Experimentally , it would be of interest to map threshold adaptiveness across different systems with a focus on directly modulating the adaptive nature , and discovering neural systems which exhibit a fixed threshold ( if they exist ) .
Mice from either sex were used according to the Guidelines of National Institutes of Health . Experiments were approved by the Institutional Animal Care and Use Committee . In vitro whole-cell current-clamp recordings were performed in acutely prepared slices of the barrel cortex after maturation of evoked neurotransmission [41] as described before [42–44] with minor modifications . Animals were anesthetized using Isoflurane before they were decapitated . Oblique thalamocortical slices [45] ( 300 mm ) were cut 45° from the midsagittal plane in chilled low-calcium , low-sodium Ringer’s solution ( in mM; sucrose , 250; KCl , 2 . 5; MgSO4⋅7H2O , 4; NaH2PO4⋅H2O , 1; HEPES , 15; D- ( + ) -glucose , 11; CaCl2 , 0 . 1 ) . Slices were incubated at 37°C for 45 minutes and kept in room temperature in carbonated ( 5% CO2 and 95% O2 ) bath solution ( pH 7 . 4 , normal Ringer’s solution: in mM , NaCl , 119; KCl , 2 . 5; MgSO4 , 1 . 3; NaH2PO4 , 1; NaHCO3 , 26 . 3; D- ( + ) -glucose , 11; CaCl2 , 2 . 5 ) . Visualized whole-cell recordings were performed using an Axoclamp-2B amplifier under an IR-DIC objective ( Olympus ) in room temperature . A bipolar extracellular stimulation electrode was placed in the lower half of a L4 barrel representing a mystacial vibrissa . A recording electrode ( 3–4 MOhm ) containing an internal solution ( pH 7 . 25; in mM; potassium gluconate , 116; KCl , 6; NaCl , 2; HEPES , 20 mM; EGTA , 0 . 5; MgATP , 4; NaGTP , 0 . 3 ) was placed orthogonal to the stimulation electrode within 150–300 μm of the cortical surface . For whole cell recordings , putative excitatory cells were selected based on pyramidal shaped somata , apical dendrites and distal tuft orientation , and regular pattern of spiking to somatic current injections ( 500 ms ) . The serial resistance Rs was compensated by bridge balance . Data was filtered ( 2 kHz ) , digitized at 5 kHz using a 12 bit National Instruments data acquisition board and acquired using Strathclyde Electrophysiology Suite for offline data analysis . All analyses were performed offline in MATLAB ( MathWorks , Inc ) . Raw voltage traces were smoothed using running window averaging ( 1 ms window size ) , and resting membrane potential ( Vm , in mV ) was calculated as the average membrane potential in a 40 ms time window prior to the stimulus onset . For those sweeps in which spike was observed , spike threshold ( Vt ) and spike latency ( St ) in respect to stimulus onset were calculated . Spike threshold was defined as the membrane potential value at which second derivative of membrane potential reached maximum as described before [7] . When analyzing neuronal responses across different resting membrane states , the data was grouped into either 3 states with 10 mV interval ( -85:10:-55 mV , Fig 7A and 7B ) or 5 states with 5 mV interval ( -82 . 5:5:-57 . 5 mV , Fig 7E and 7G ) based on the Vm . We employed the Shannon information theory [51] to analyze the stimulus representation capacity of neurons with either adaptive or fixed spike threshold . The mutual information MI between the stimulus S and neuronal response R is calculated as I ( S , R ) =H ( R ) −H ( R|S ) in which H ( R ) denotes the entropy of the response variable R: H ( R ) =−∑i=1np ( ri ) log2 ( p ( ri ) ) and H ( R|S ) is given as H ( R|S ) =−∑i=1np ( si ) ∑j=1mp ( rj|si ) log2 ( p ( rj|si ) ) where m , n is the number of possible response and stimulus patterns and p ( r ) and p ( s ) is the occurrence probability of these patterns , respectively . Intuitively , the mutual information measures how much uncertainty about one variable , either R or S , can be reduced by an ideal observer observing the other variable . It is theoretically the upper bound of knowledge can be gained from the neuronal response under a given condition ( i . e . no other source of additional information ) . To construct response patterns , we binned the neuronal response spike train from individual trials using various temporal bin sizes ( 1 ms , 2 ms , 5 ms or 30 ms ) . The analysis window was 0–30 ms after stimulus onset , which included all spikes elicited by the stimulus . Number of spikes in each temporal bin was counted , and the resulting numeric vector with different length , depending on temporal bin size , was used to calculate the mutual information . When calculating the mutual information between stimulus and neural response knowing the states ( Figs 5–7 ) , the states was supplied to the binned response vector as an additional dimension . The mutual information calculation was performed using Spike Train Analysis toolbox [52] with shuffle correction combined with Panzeri-Treves estimator [53] as bias correction method [54] . With the bias correction method employed , the number of trials per stimulus ( Ns ) should be at least ¼ of the number of total possible response patterns ( NR ) to obtain an accurate estimation of the mutual information [54] . In our simulations a neuron fires at most 2 spikes in each trial . Using a 30 ms analysis window and 2 ms bin size , the NR would be ( 152 ) + ( 151 ) + ( 150 ) =121 , thus the minimum Ns required is 31 . If a bin size of 1 ms is used , then NR = 466 and the minimum Ns would be 117 . With Ns = 150 in our dataset , the calculated mutual information should be bias-free . We ran a set of simulations with Ns = 1000 and compared the MI calculated using Ns = 1000 and those calculated with various smaller Ns using bootstrapping method , setting analysis window to 30 ms and bin size to 1 ms . At Ns = 150 , the difference is already <0 . 1% ( <4x10-3 bit in absolute value ) of the true MI . The total information that can be decoded without knowing the state is given by I ( S; R ) , which is best interpreted as the information robust to state changes , i . e . the information can be decoded without paying attention to state . The total information that can be decoded knowing the state in addition , is given by I ( S; R , states ) . We calculated the robustness index RI , which is defined as I ( S; R ) /I ( S; R , states ) . RI values close to 1 imply a high robustness of the decodable information to not knowing states , i . e . most of the total information can be obtained without the state knowledge . Conversely , low RI values indicate that taking state into account significantly adds to the amount of information the postsynaptic neuron transmits . Since RI is relative to I ( S; R , states ) and both constituents are positive , RI is bounded in [0 , 1] . The decoding with moving reference time was implemented by taking the average spike time of all simulated neurons ( N = 500 ) in response to the stimulus . Assuming a flat distribution of tuning preferences , this average time will even be stimulus independent . The early part of the decoding window ( set of discretization bins ) was then aligned to the average response time ( similar to [25] ) . Note , that the neural responses could differ by state and thus also shift the average response time forward or backward . | A neuron is a tiny computer that transforms electrical inputs into electrical outputs . While neurons have been investigated and modeled for many decades , some aspects remain elusive . Recently , it was demonstrated that the membrane ( voltage ) state of a neuron determines its threshold to spiking . In the present study we asked , what are the consequences of this dependence for the computation the neuron performs . We find that this so called adaptive threshold allows neurons to be more focused on inputs which arrive close in time with other inputs . Also , it allows neurons to represent their information more robustly , such that a readout of their activity is less influenced by the state the brain is in . The present use of information theory provides a solid foundation for these results . We obtained the results primarily in detailed simulations , but performed neural recordings to verify these properties in real neurons . In summary , an adaptive spiking threshold allows neurons to specifically compute robustly with a focus on tight temporal correlations in their input . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"action",
"potentials",
"medicine",
"and",
"health",
"sciences",
"neural",
"networks",
"depolarization",
"membrane",
"potential",
"electrophysiology",
"neuroscience",
"simulation",
"and",
"modeling",
"computational",
"neuroscience",
"research",
"and",
"analysis",
"methods",
"excitatory",
"postsynaptic",
"potentials",
"computer",
"and",
"information",
"sciences",
"animal",
"cells",
"cellular",
"neuroscience",
"cell",
"biology",
"physiology",
"neurons",
"single",
"neuron",
"function",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"computational",
"biology",
"neurophysiology"
]
| 2016 | Adaptive Spike Threshold Enables Robust and Temporally Precise Neuronal Encoding |
Thousands of loci in the human and mouse genomes give rise to circular RNA transcripts; at many of these loci , the predominant RNA isoform is a circle . Using an improved computational approach for circular RNA identification , we found widespread circular RNA expression in Drosophila melanogaster and estimate that in humans , circular RNA may account for 1% as many molecules as poly ( A ) RNA . Analysis of data from the ENCODE consortium revealed that the repertoire of genes expressing circular RNA , the ratio of circular to linear transcripts for each gene , and even the pattern of splice isoforms of circular RNAs from each gene were cell-type specific . These results suggest that biogenesis of circular RNA is an integral , conserved , and regulated feature of the gene expression program .
Recently , we were surprised to find that the predominant RNA isoform from hundreds of human genes is a circle , and that circular RNAs were transcribed from thousands of genes in both human and mouse [1] . Circular RNA transcripts had been reported previously for a handful of genes . With the possible exceptions of the circular RNA isoforms of the Sry gene in mouse testis [2] and the muscleblind gene in Drosophila melanogaster [3] these were generally thought to be rare RNA species , perhaps the result of transcriptional noise . In humans , circular isoforms of the transcripts from the ETS and cytochrome P450 2C24 genes have also been reported; these were found to be inabundant compared to linear RNA isoforms from the same genes [3]–[5] . In recent years , two antisense circular RNAs were discovered and studied more intensely in humans: an antisense transcript from the INK4A-ARF locus , cANRIL , and an abundant antisense transcript to CDR1; the latter was recently reported to be a microRNA sink [6]–[9] . The ubiquitous expression of circular RNA in human and mouse cells has now been independently confirmed by high throughput sequencing of the RNase R treated , ribosomal-depleted fraction of RNA , combined with a previously published informatic algorithm to identify circular RNA [7] as well as by a second report characterizing RNA after ribosomal RNA depletion [8] . In the former report , a large majority of the circular isoforms we had described ( 1025 of 1319 ) were also identified by deep sequencing of RNase R-treated RNA . This overlap in circles identified in these two studies is noteworthy because the more recent report focused on fibroblasts , while we previously analyzed RNA isolated from leukocytes and pediatric leukemias . Here , we describe a more systematic bioinformatic and statistical genome-wide study that significantly expands the catalogue of circular RNAs identified in human cells and reveals significant regulation of circular RNA expression . In many applications , computational challenges associated with mapping and with distinguishing between sequencing errors and sequence homology prevent reliable identification of structural variants , including circular RNAs . Indeed , although de novo splicing detection algorithms have been used in more than a thousand published studies , including studies aimed at identifying gene fusions and internal tandem duplications , most instances of scrambled exons in human RNAs , and thus the circular species that they represent , had eluded detection .
A major challenge in bioinformatic and statistical identification of novel RNA isoforms , particularly circular RNA , involves distinguishing bona fide evidence of scrambled exons in RNA from confounding factors such as sequence degeneracy at exon boundaries and sequencing errors . To address these challenges and to identify circular isoforms from public ENCODE RNA-Seq data , we developed a new bioinformatic approach . The main idea behind our computational method is that it refrains from qualitative hard thresholding of read alignment quality , and instead computes statistical averages of alignment quality scores . This approach allows us to distinguish putative novel splice junctions where the majority of reads align to the ‘novel’ junction with high quality alignment scores from those where reads with high alignment scores are rare . The method allows for systematic FDR-based thresholding , rather than qualitative cut-offs , to determine classification as a scrambled junction at a prescribed confidence level . Our method was focused on identifying circular RNA transcribed from genes whose linear isoform exons are annotated: we first built a database of all scrambled junctions between annotated exon boundaries , essentially as previously described [1] , extending the database to annotated hg19 UCSC ‘knowngene’ exon boundaries . Importantly , we did not impose a lower threshold on the length of annotated exons in our database , instead generating database entries of short annotated exons by an ‘in silico’ rolling circle . We required a minimum of 10 nt on both sides of a diagnostic read to span a scrambled exon-exon junction . The improved sensitivity of this approach compared to previous methods allowed us to identify thousands of previously unreported circular isoforms and some very small circular RNAs , exemplified by a <150 nt circular RNA isoform ABTB1 resulting from the splicing of two short exons . Other small circular isoforms that we identified and confirmed include a 204 nt circular isoform from a single exon of LINC00340 - a long intergenic noncoding RNA , and a two exon circle of 151 nt from the RNA binding motif gene RBM5 . Experimental and bioinformatic noise can give rise to spurious evidence of circular transcripts , especially for highly expressed genes . To tackle this problem , we combined the bioinformatic approach above with a statistical strategy to distinguish reads supporting exon scrambling from reads likely to be homology and sequencing artifacts . Briefly , we did not impose any thresholds on alignment quality of either read 1 ( aligning to a diagnostic scrambled exon-exon junction ) or read 2 ( aligning to a canonical isoform or a diagnostic scrambled junction ) . All read pairs with evidence of scrambled exon splicing at a particular pair of genomic coordinates were aggregated by averaging , measuring the overall quality of reads aligning to the putative circle . We generated an empirical null distribution of average alignment qualities using “decoy” read pairs . These “decoy” reads had the property that read 1 mapped to a scrambled intragenic exon X – exon Y junction and read 2 mapped within the same gene but would be excluded from a circle composed of exons Y , Y+1 , … X , see Figure 1A . These alignment qualities were averaged across all reads for each circular RNA and generated our null distribution of the alignment quality as depicted in Figure 1B . This approach allowed us to compute a per-isoform FDR ( by referring the alignment score per isoform to the empirical null distribution ) and reduce calls of false positive circular isoforms which riddled the data before this approach was applied . Using this approach , we were able to enhance statistical discrimination between case 1: false positive evidence of circular RNA isoforms in highly expressed genes resulting from reads with sequencing errors observed due to high sampling of these genes , and case 2: bona fide low-level expression of circular isoforms from these highly expressed genes . For example , the vast majority of reads ( 99 . 995% ) from putative circular isoforms of GAPDH had a FDR significantly surpassing our threshold of . 025; these reads would be taken as evidence of circular isoform expression with a naïve approach . We applied our method to a large publicly available data set from the ENCODE consortium ( see Table 1 ) , with the goal of identifying novel RNA circular isoforms and studying regulation of circular RNA . This dataset consisted of 76-nt paired-end reads from RNA isolated from 15 different cancer and non-cancer cell lines representing mesodermal , ectodermal , endodermal lineages and the pluripotent H1-hESC ( see Table 1 ) . Each RNA sample was depleted of poly ( A ) RNA , size selected to be above 200 nt , and subsequently subjected to ribosomal RNA depletion by RiboMinus ( see ENCODE protocols ) . We have also made use of public analysis of the matched poly ( A ) -selected samples from these cell lines published by the consortium . Note that the statistics above are absolute counts not corrected for sequencing depth , which varied by sample . Across the 15 cell types , at an FDR of . 025 , we found 46866 distinct intragenic splice junctions at annotated exon boundaries in 8466 genes . Across cell types , we detected the largest number of genes with evidence of circular RNA expression in the leukemia cell line K562 ( 16559 distinct circle-specific splice junctions ) ; in the fetal lung fibroblast line AG04450 , we detected 11590 distinct splicing circle-specific splice junctions and in the human foreskin fibroblast line BJ , we detected 7771 ( this is not a typo: [7] reports exactly the same number of circular isoforms ) . Recently , 7771 , 2229 and 485 splice junctions , of ‘low’ , ‘medium’ and ‘high’ stringency , respectively , representing circular RNA were identified by another method in Hs68 cells , a human fetal foreskin fibroblast line [7] . We used the enzyme RNase R , a highly processive 3′ to 5′ exoribonuclease , to test our computational predictions of circular RNA species . This exonuclease is not expected to digest circular RNA because they lack the required free 3′ end but readily digests linear RNAs with a 3′ single stranded region of greater than 7 nucleotides [10] . We tested a panel of 8 putative circular RNAs varying in size , abundance and abundance of the corresponding linear isoforms: ABTB1 , FAT1 , HIPK3 , CYP24A1 , LINC00340 , LPAR1 , and PVT1 . As positive controls , we included 3 genes with strong prior evidence of circularization: MAN1A2 , RNF220 and CAMSAP1 . We treated total RNA from HeLa cells with either RNase R or a mock enzyme treatment . For each sample , we performed an RT with random hexamer primers and used qPCR to quantify the change in abundance of species with scrambled exons compared to species with exons that we predicted found only in linear RNA isoforms , following treatment with RNase R . All the RNA species that we had predicted to be circular were resistant to RNase R whereas all predicted linear sequences were highly sensitive to RNase R ( see Figure 2 ) , providing strong evidence that our computational method specifically identifies circular RNA species . Because the ENCODE libraries were constructed to preserve strand directionality , we could analyze the data for evidence that circular RNAs may serve as an RNA-dependent RNA polymerase as has been shown to occur in some siRNA pathways [11] and in viral or viroid replication [12] . Among paired end reads that support either sense or antisense circular RNA , with a diagnostic circular junction between exon boundaries annotated for linear RNA , we found a strong and significant bias in the directionality of reads ( almost 100% of reads from >93% of putative circles ) . This bias supports the hypothesis that the significant majority of RNA circles formed using splice sites shared with annotated linear RNAs are transcribed from the same strand as the canonical linear RNA . By this analysis , the percentage of circular isoforms in the sense orientation ( with respect to the linear isoform ) was 96% for HMEC and >99% for the other 14 cell types . This evidence argues against a primary function of circular RNA serving as an RNA template for an RNA-directed polymerase . A small minority of reads had a polarity inconsistent with transcription in the same direction as linear RNA; we believe in most cases these represent artifacts of reverse transcription , perhaps induced by RNA secondary structure . Note that our intention was not to identify un-annotated antisense circular RNA , and we did not search for circular RNAs that might have been produced by splicing a primary transcript complementary to the annotated transcript . We sought to determine the relative abundance of each circular RNA compared to its cognate linear RNA . This required estimating the relative abundance of each linear RNA , the relative abundance of each circular RNA , and an “equivalence factor” or normalization constant relating the number of RNA molecules represented by 1 measured unit of linear RNA to 1 measured unit of circular RNA . For linear RNA abundance , we used the estimates generated by the public ENCODE consortium analysis of polyadenylated fractions , represented in RPKM units ( reads per kilobase of transcript per million mapped reads in the sample ) . Our estimate of each circular RNA isoform's abundance from sequencing was the number of read pairs , in poly ( A ) -minus fractions , in which one read spanned a circular junction ( note that counting junctional reads in this manner inherently normalizes by gene length ) . To determine the equivalence factor ( the number of junctional read counts expected for a circular RNA expressed at the same level as a transcript with an RPKM of 1 ) , we measured the abundance of circular and linear isoforms of FAT1 and HIPK3 across three ENCODE cell lines ( A549 , AG04450 and HeLa ) by qPCR . This allowed us to relate the abundance of the linear isoforms of FAT1 and HIPK3 as measured in units of RPKM to the abundance of the circular isoforms of these genes as measured in units of junctional read counts . Since the equivalence factor is the same for all genes in the genome , we were then able to compute the relative abundance of circular and linear isoforms for all genes detected in the sequencing data . FAT1 and HIPK3 were chosen because they have large , abundant circular RNA isoforms and high linear RNA isoform expression , thus mitigating potential factors confounding this estimation such as rolling circle amplification of small circular RNA during the RT , and statistical uncertainty introduced by estimating the expression of low abundance circle or linear isoforms . These estimates suggested that there was roughly 1 molecule of circular RNA for every 100 molecules of poly ( A ) RNA in the cell lines we evaluated: A549 , AG04450 and HeLa . For roughly 50 genes in each cell line , circular transcript isoforms were estimated to be more abundant than linear isoforms ( see Tables S1 , S2 , S3 for a complete list of the relative linear: circular isoforms per gene genome-wide ) . For most genes with circular RNA isoforms , the abundance of the circles was roughly 5–10% that of their linear counterparts . At least among this small sample of cell lines , the differences in growth rate and developmental origin do not appear to fundamentally alter the genome-wide rate of circular RNA expression . As a spot-check of our sequencing based estimates of relative abundance of linear and circular isoforms , we performed a Northern blot for CAMSAP1 with total RNA from HeLa cells ( see Figure 3 ) . Sequencing based estimates suggested that the circular isoform of CAMSAP1 consisting of exons 2 and 3 was 20 times more abundant than the linear counterpart . The Northern blot shows that CAMSAP1 circular isoforms are more abundant than the linear isoform . Intriguingly , one of the major bands ( at 1446 nt ) is an unexpected circular isoform consisting of exon 2 - intron 2 - exon 3 . RT-PCR bands consistent with both isoforms were amplified from RNase R treated HeLa RNA; Sanger sequencing of the gel-purified bands verified their structure ( data not shown ) . RNA-Seq also provided evidence for this retained-intron circle: in poly ( A ) depleted fractions of HeLa-S3 , the highest read density was in exons 2 and 3 followed by intron 2 , with other introns more than 10-fold lower . Of note , our estimate of the ratio of intron 2 to exon 3 expression ( based on number of reads with zero mismatches to the genome ) was somewhat higher in the nuclear fraction ( 38% ) compared to in the cytosolic ( 10% ) or RNA isolated from total cells ( 16% ) . There was also cell type variation of the ratio of intron 2 to exon 3 read density in “cell” fractions across the ENCODE data set , from 16% in HeLa-S3 down to 3 . 5% in NHEK , suggesting that intron-retention in CAMSAP1 circles may be under regulatory control . We previously reported that genes with circular RNA transcripts tend to have larger introns than genome-wide averages . That analysis showed that even after controlling for the tendency for intron lengths to decrease from 5′ to 3′ along the canonical transcript [1] , [13] , the introns immediately flanking the exon boundaries that participated in the scrambled splice were significantly longer than average . Here we further investigated the relationship between intron length and circularized exons in this deeper survey of circular RNA expression . For each UCSC annotated gene , for each annotated splice site , we defined the flanking intron length for the 3′ and 5′ splice sites as the distance to the nearest upstream or downstream 5′ or 3′ splice site respectively . To control for systematic biases , for example , that genes expressing circular RNA isoforms have relatively large introns compared to genome-wide averages ( as we have found previously ) , we performed the following analysis . We ranked introns that flanked spliced exons generating circular RNA isoforms in two ways: 1 ) weighting the lengths of introns flanking each circular isoform by the abundance of the corresponding circular RNA ( right panel in Figure 4 ) ; 2 ) counting each circular isoform once regardless of its expression level ( left panel in Figure 4 ) . For the analysis depicted in Figure 4 , for each gene , we ranked the length of each intron according to its length . We then converted each rank value to a quantile: for example , the second largest intron in a gene with 5 introns would receive a quantile of 40 ( = 2/5 * 100% ) . For reference , under a null model where the rank of intron length had no relationship with propensity to flank a circular splice donor or acceptor , the heatmaps would have uniform intensity regardless of the quantile represented . We found that the relative length of the flanking intron did not reliably determine which exons were spliced to form an RNA circle , although within a gene , longer introns were more likely to flank circularized exons . To test whether small variations in intron length might explain the dynamic range in intron length quantiles observed in Figure 4 , we also examined the relative length of each intron flanking a diagnostic donor or acceptor site in the circle as a fraction of the largest intron length in the gene ( Figure S1 ) . Thus , if one of the introns flanking a splice site diagnostic of a circle were the longest intron in the gene , its ratio compared to the maximum intron length would be 1 . The null distribution we considered was based on the relative length of the second vs . third largest intron in the set of genes evaluated . Unexpectedly , we found that , measured as a fraction of maximum intron length , introns flanking circular junctions were , on average , smaller than expected from the null distribution , perhaps explained by a single atypically long intron within genes expressing circular isoforms . We explored regulation of circular RNA expression using the ENCODE RNA-Seq data for a number of cultured cell lines , then did an independent evaluation of some of the identified circle expression variation using qPCR ( see Figure 5 ) . Some of the genes we tested by qPCR ( CYP24A1 , PVT1 and LPAR1 and LINC00340 ) expressed circular RNA isoforms that were predicted from sequence data to vary across cell lines; others ( FAT1 , HIPK3 ) appeared from the RNA-Seq data to have constant levels of circular isoform expression in A549 , AG04450 and HeLa cells ( data not shown ) . To assess variation in circular RNA expression genome-wide , we estimated the abundance of circular RNA from sequence data based on the diagnostic splice-junction counts described earlier; estimates of linear transcript abundance were from the ENCODE consortium's analysis of poly ( A ) gene expression . For a set of relatively highly expressed circular isoforms , we evaluated the fit of a Poisson model in which circular RNA expression increased with linear isoform abundance , controlling for effects of sequencing depth and incorporating experimental variation by treating experimental replicates as distinct . We also included cell-type effects to further account for circular RNA expression . We observed the largest dynamic range in circular RNA production in the gene CYP24A1 , a candidate oncogene encoding a component of the vitamin D3 metabolic pathway . Its linear mRNA product and the CYP24A1 protein have been reported to be expressed at elevated levels in many primary lung and other cancers , and in many lung cancer cell lines , including A549; no amplification of the CYP24A1 gene has been reported in A549 [14]–[19] , although CYP24A1 is frequently amplified and mutated in primary human cancers [20] . Our statistical model also suggested that other highly expressed RNA circles had cell-type specific increases in expression that could not be accounted for by an increase in overall expression of the corresponding linear RNA . One example is the circular isoform of DOCK1 , whose linear isoform encodes a “dedicator of cytokinesis” , a RacGEF , and was the most highly expressed circular isoform in MCF-7 , a breast cancer cell line . DOCK1 also had the highest estimated ratio of circular: linear RNA expression in MCF-7 among all cell types in the ENCODE panel , including those where the linear isoform of DOCK1 was more highly expressed ( see Figure 6 ) . Figure 6 depicts other examples of genes with cell-type-specific selective increases in the ratio of circular to linear RNA isoforms . One example was the much higher expression of a circular RNA isoform of RBM33 in K562 cells compared to the other cell types . We have previously detected RBM33 circles in human leukocyte and leukemia samples and mouse brain [1] , suggesting the possibility of evolutionary conservation . RBM33 has not been extensively studied , but duplication of a locus including Sonic Hedgehog and RBM33 has been associated with congenital muscular hypertrophy [21] . Similarly , expression of a circular isoform of the long intergenic noncoding RNA LINC00340 was specifically elevated in H1-hESCs . In H1-hESCs , sequencing data suggested that the circular isoform of LINC00340 was the fourth most highly expressed circular RNA of all detected circular isoforms . Circular isoforms of other LINC RNAs , including LINC00263 and LINC00265 , were also identified in our analysis . Because noncoding RNAs , including LINC RNAs are generally less well annotated than messenger RNAs , it is possible that our analysis may still have under-detected circular isoforms of these RNAs as we did not specifically attempt to improve their representation in the UCSC knowngene annotation . A final highlighted example of cell-type-specific selective increases in the ratio of circular to linear RNA isoforms in Figure 6 is AMBRA1 . Two different circular isoforms of AMBRA1 RNA were differentially regulated in MCF-7 and HepG2 cells . MCF-7 cells expressed higher levels of a longer isoform ( 362 nt ) while a shorter isoform ( 182 nt ) was more highly expressed in HepG2 . AMBRA1 plays a key role in autophagy; deficient mice have excessive cell death by apoptosis [22]–[23] . In these specific examples , and in general , variation in the abundance of hundreds of circular RNA isoforms appeared to have little or no correlation with variation in the abundance of the cognate linear RNA between the cell lines we compared . In particular , we did not observe a correlation between circle-specific junctional counts and overall abundance of the corresponding RNA as measured by RPKM , even at the lowest levels of gene expression . Further evidence that RNA circles are not just an accidental aberration of normal splicing is provided by the fact that circular RNA isoforms are generated by splicing very specific pairs of exons ( see discussion below ) . When a gene encodes multiple alternatively spliced circular isoforms , what patterns characterize the use of splice acceptor and donor pairs to produce the circle junction ? To study this question , we distinguished three broad classes of splice site pairings , which we term stereotyped , proximal and combinatorial pairing , respectively . Examples of each are depicted in Figure 7 . For most genes that have circular RNA isoforms ( the “stereotyped” class ) , a single splice site donor and acceptor pair were either used exclusively or strongly preferred in the splice that gave rise to the circular isoforms of the gene; this was the case , for example , for CYP24A1 and MCU . While CYP24A1 was the most highly expressed circular RNA among the examined cell lines and MCU was among the 20 most highly expressed circular RNAs in 9 different cell types , only one circular splice variant from each gene was overwhelmingly preferred ( see Figure 7 ) . A variant of stereotyped splicing was exemplified by the circular RNA isoforms of MBOAT2 . These isoforms were expressed at levels similar to MCU , but with a distinctly different pattern of splicing: one particular splice acceptor was highly preferred , but several alternative splice donors were used and each produced similar levels of the corresponding spliced circular RNA isoform . It is noteworthy that none of the exons that participate producing the MBOAT2 circles have been reported to participate in alternative splicing of linear RNA MBOAT2 isoforms . For many transcripts in which multiple splice donor and multiple splice acceptor sites were used in circular splicing , proximal donor-acceptor pairs were strongly preferred . This “proximal” pattern of circular splicing is exemplified by the circular isoforms of ABCC1 . The “combinatorial” pattern of circular splicing is exemplified by CAMSAP1 and especially PICALM . Multiple splice donors and multiple splice acceptors participate in production of circular isoforms , with little preference for proximal donor and acceptor sites . In contrast to PICALM , across cell types , CAMSAP1 has a single dominant isoform . Although detection of rare circular RNA isoforms increased with sampling depth of the RNA sequences ( see Figure S3 ) , within a gene , our data did not fit a simple model where overall expression of circular RNA isoforms predicted the diversity of circular RNA isoforms expressed ( Figure S4 ) . Figure S4 depicts other features of intragenic circular RNA splicing patterns across all genes: the majority of genes with detectable circular RNA expression had detectable levels of more than one circular isoform . Also in such genes , the number of splice donor and acceptor sites used in circular splicing was correlated: when more acceptor sites were used in circular RNA products from a particular gene , so were more donor sites . Considering all genes with circular RNA isoforms , we found that cells generally expressed a small fraction of the number of circular RNA isoforms that could , in principle , be produced by splicing a downstream splice donor to an upstream splice acceptor ( see Figure S4C ) . We quantified this fraction by defining a combinatorial index C , which compares the number of observed circular isoforms to the number of possible pairings of the detected acceptor and donor splice sites ( see Methods ) . In general , regardless of the total expression level of circular RNA isoforms , half or less of the combinatorial space of circular isoforms—conditioned on acceptor donor and acceptor sites used in circular RNA splicing– had detectable expression , and many genes expressed the minimum number of potential circular RNA isoforms represented by the lowest value of C . We used a statistical model to identify genes with regulated use of donor and acceptor sites characterizing the diagnostic non-canonical exon junction . For each gene and each cell type , we estimated the frequency with which each donor and acceptor splice site was used ( see Tables S4 , S5 ) , and computed confidence intervals for the use of each site by cell type . For hundreds of genes , we found statistical evidence of cell type-specific preferences in patterns of splice donor and acceptor usage in the biogenesis of circular RNA ( Tables S4 , S5 ) . Three of these genes are shown in Figure 8 . The RNF19B gene provides a simple and striking example . The only circular isoform of RNF19B RNA detected in NHLF was undetectable in any of the other cell lines examined . Conversely , the dominant circular RNF19B isoform in the other cells was undetectable in NHLF ( see Figure 8 ) . In a second example , a single splice acceptor was used in all circular LPAR1 RNAs identified in NHEK , NHLF and HSMM cells , whereas three different splice acceptors were represented in the circular LPAR1 RNAs found in two fetal fibroblast cell lines , AG04450 and BJ . The differences in diversity of circular isoforms were not explained by cell-type specific differences in overall LPAR1 expression . ZFAND6 is a third example of a gene with regulated circular RNA expression . A549 cells expressed a single circular isoform , while the remaining cell types expressed two circular isoforms . These differences cannot be readily explained either by differences in sequencing depth , cell-type-specific differences in linear or circular RNA expression or any cell-type independent differences in the RNA , such as intron lengths or structure ( see Figure 8 ) . For example , among all the cells we examined , NHLF expressed the second highest levels of linear ZFAND6 RNA , but circular ZFAND6 RNAs were undetectable in these cells . Further , we do not observe any correlation between canonical alternative splicing and circular RNA splice site use or patterns in the three genes depicted in Figure 8 ( see Figure S2 ) . To further assess evolutionary conservation of circular RNA expression across model organisms , we surveyed circular RNA expression using published RNA-Seq data from Drosophila brains [24] . This analysis revealed hundreds of genes encoding circular RNA isoforms in fly , including abundant expression of a previously described circular RNA isoform from the muscleblind locus [3] . Muscleblind was among the most highly expressed circular isoforms , but our analysis indicated that circular RNAs from 11 other genes in these samples were even more abundant: the most highly expressed putative circular RNAs were from Pka-C3 , encoding a cAMP-dependent protein kinase , and scarecrow ( scro ) , encoding an NK-2 homeobox protein . Other highly expressed RNA circles included Caps , ps , mGluRA , caps , snap25 , jp , zfh2 and two genes of unknown function , CG40178 and CG17471 . Overall , we found evidence for exon scrambling in more than 800 distinct Drosophila splice junctions supported by more than one sequencing read ( Table S6 ) . Additional evidence supporting some evolutionary conservation of circular RNAs is found by considering mouse genes represented in brain RNA-Seq data [1] . Genes whose human orthologs expressed circular RNAs were statisteically more likely to have evidence for circular isoforms in the mouse RNA-Seq data . Roughly 4% of genes with expression in both mouse and human data and which encoded orthologous proteins also encoded circular RNA detected in both data sets compared to a null expected rate of 2 . 5% . This suggests modest conservation of circular RNA expression from loci with orthologous protein products , ignoring finer features that might influence conserved expression of circular RNA , such as developmental stage . In addition , several genes encoding exclusively non-coding RNA species , including IPW ( Imprinted in Prader-Willi syndrome ) and the oncogene PVT1 were expressed as circular isoforms in both mice and humans .
Characteristic changes during development and differentiation are a pervasive feature of global gene expression programs . We systematically searched for evidence of circular RNAs in a large corpus of RNA-Seq data generated by the ENCODE consortium as well as in RNA-Seq data from Drosophila brain . We found that circular RNA comprises a significant fraction of cellular RNA and that the relative abundance of circular isoforms and the specific patterns of splice site usage in RNA circularization are regulated in a gene-specific and cell-type specific manner . The results strongly suggest that circular RNAs are a common , abundant and potentially developmentally regulated component of the gene expression programs in diverse animal species . To improve our sensitivity and specificity in detecting circular isoforms , we developed improved bioinformatic and statistical methods that enabled more reliable discrimination between bona fide evidence of exon scrambling and artifacts introduced by high throughput sequencing and sequence homology within a gene . This improved performance allowed us to detect a more extensive catalog of circular RNA than previously reported , including small RNA circles , RNA circles formed by non-canonical splicing of short exons and noncoding RNAs . Improved detection of circular RNA isoforms has also allowed us to characterize the extent of differential circular RNA splicing within a single gene , and to study variation in alternative splicing of circular RNA; indeed , this method may have wider applicability in the discovery of novel RNA splice junctions and detection of other variant sequences . qPCR quantification and extensive analysis of RNA-Seq data has allowed us to estimate that the number of circular RNA molecules is roughly 1% of the number of poly ( A ) molecules in the cells we investigated . This estimate is remarkably similar to a report published more than 30 years ago , which found physical evidence of circular RNA by examining cellular RNA by electron microscopy [25] . We tested the hypothesis that circular RNAs might be the result of a background “noise” level of dysfunctional splicing . Under this model , we would expect a positive relationship between linear RNA isoform expression from a given gene and the probability of detecting a circular RNA isoform from that gene . We found no evidence of such a relationship , suggesting instead that even low rates of circular RNA production are regulated , or that highly expressed genes have evolved specific mechanisms to prevent splicing errors that could result in RNA circles . Our initial report of the ubiquity of circular RNA , based on sequencing an ribosomal-RNA depleted RNA fraction , has since been confirmed in an independent study in which circular RNAs from human and mouse fibroblasts were enriched by treating RNA with RNase R , and in a second genome-wide search for evidence of circular RNA by sequencing ribosomal-RNA-depleted RNA samples [7] , [8] . The analysis presented here significantly expands the catalog of circular RNAs expressed by humans and Drosophila . It is likely that human cells express even more circular RNAs than we report here: we did not attempt a ‘de novo’ identification of circular RNA , and instead focused on circular RNA produced by splicing at annotated exon boundaries . For example , by analysis of a Northern blot for CAMSAP1 in HeLa cells , and a subsequent limited bioinformatic survey of 6 genes , we found evidence of cell-type specific variation in rates of intron retention as well as evidence that circular , intron-retained transcripts in HeLa-S3 cells may be sequestered in the nucleus and potentially exported to the cytoplasm . CAMSAP1 is a calmodulin regulated gene and has conserved circle expression in mouse and Drosophila ( spp4 ) , and it would be interesting to study intron retention in these organisms . The most highly expressed circular RNA identified in our analysis was from the CYP24A1 gene , in a lung cancer cell line , A549 . CYP24A1 , which encodes 1 , 25-dihydroxyvitamin D 3 24-hydroxylase , has been suggested to play a role in the pathogenesis of many carcinomas [14] , [18] , [26]–[32] . We found that the circular isoform was expressed at levels comparable to the canonical linear form . Although the circular isoform of CYP24A1 ( which includes all but the first and last exons of CYP24A1 ) could in principle encode a protein lacking the N terminal mitochondrial localization signal , we have not found evidence for such a protein by mass spectrometry on A549 cell lysates ( unpublished data ) . This finding is consistent with other evidence that despite the formal possibility of translation of circular RNA by the eukaryotic ribosome [33]–[34] , circular RNAs do not in general act by encoding a protein . Recent reports have shown that an antisense circular transcript from the CDR1 locus is enriched for functional microRNA binding sites [7] , [8] . However , in a preliminary analysis , we have not found evidence that enrichment of microRNA binding sites is a global feature under selection in the sequence of the thousands of circular RNAs profiled in this paper . For example , we see a roughly 5% enrichment of microRNA binding sites in a 66 nt window in sequence flanking diagnostic circular RNA junctions in circular RNAs which are highly expressed in at least one cell type compared to the number of binding sites in the junctional sequences flanking all detected circular RNA junctions ( see Figure S5 ) . The low enrichment is perhaps not surprising considering that the vast majority of the transcripts we surveyed in [1] and in this report are transcribed in the same sense with respect to the linear mRNA isoform . Therefore , except for ‘diagnostic’ junctional sequence and secondary and higher order structure , circular and linear isoforms would have the same potential to bind microRNA , albeit with a different degree of stability . Certainly , the potential genome-wide interplay between microRNAs and circular RNAs warrants further experimental and computational investigation . Our findings that mouse orthologs of human genes with circular RNA products are themselves more likely to encode circular RNAs are consistent with a similar independent analysis of circular RNA conservation and support the hypothesis that circular RNAs have an evolutionarily conserved function [7] . Thus , although the abundance , ubiquity , and potential developmental regulation of circular RNAs all point to the possibility of important functional roles , their nature and mechanisms are still to be discovered .
Raw fastq files available on Sept . 3 , 2012 were downloaded from the ENCODE project website and processed in batch using custom Perl scripts . At that time , 2 replicates from each of 15 cell types were available , with the exceptions that 1 HMEC and 3 NHEK were downloadable . We selected all long poly ( A ) minus reads banked at http://hgdownload . cse . UCSC . edu/goldenPath/hg19/encodeDCC/wgEncodeCshlLongRnaSeq . Read 1 and Read 2 reflect directionality of original RNA and were not processed symmetrically . We constructed a custom database for sequence alignment as follows: all UCSC annotated exons in scrambled order were identified and for each pair , 66 nt from each side the 3′ and 5′ ends of flanking exon were concatenated . Sequence alignment of a 76 nt read hence required alignment with a minimum of 10 nt overhang . Cases where exons were <66 nt were treated separately , by concatenating exons upstream of the donor or downstream of the acceptor exon in the scrambled pair , or using ‘in silico’ rolling circle concatenation in cases where the total circle size was smaller than 132 nt . In detail , read 1 and read 2 were not treated symmetrically as the input library was a directional RNA-Seq library . Read 1 was aligned to UCSC knowngenes and the human genome under bowtie2 default conditions [35] . Reads failing this alignment were aligned to a custom database of all scrambled exon-exon junctions for each UCSC knowngene isoform , again under bowtie2 default conditions . We culled the mate of each aligned read 1 , and performed an alignment of this subset of reads to the above UCSC knowngenes and to the custom database of scrambled exon-exon junctions: ( thus , in principle , we could have analyzed the data as described above focusing on read 2 and increased the number of detected junctional reads ) . In conjunction with the alignment to the above database of exon-exon junctions , we modeled the null distribution for rates of mismatch of reads aligning to this database using a method that should be of general interest in discovery of structural variants using high throughput sequencing data . In overview , we considered all reads that aligned with qualities a ) and b ) below , without imposing hard thresholding on the quality of alignment of either read: A read reflecting a circular RNA isoform transcribed from the same strand as the canonical isoform has the property that read 1 maps to the −orientation and read 2 to the +orientation . Alignment scores were calculated using the bowtie2 default which for example , adds a ‘−6’ penalty for a mismatch between the reference and aligned read at a high quality base call . Summing penalties for mismatches produces an overall alignment score per read , one score for the read spanning junction ( read 1 ) and one score for read 2 . Three statistics measuring alignment were calculated for each pair of scrambled exons for each UCSC isoform supported by at least one read: the average alignment score of read 1 , the average alignment score of read 2 , and the average product alignment score of read 1 , read 2 although this measure was not ultimately used to calculate the FDR reported . In detail , to compute the FDR , we created a null distribution of the joint alignment statistics for read 1 and read 2 using reads where read 1 mapped to a junction between exon x and exon y ( x> = y ) and read 2 mapped upstream of exon y or downstream of exon x which is incompatible with it deriving from a circular RNA molecule . We used the pair of ( read 1 , read 2 ) alignment statistics from such reads to generate the FDR per isoform depicted in Figure 1 . Subsequently , all reads were filtered to an FDR level of . 025 unless otherwise specified . See Table S7 for a complete list of scores . HeLa total RNA was isolated by TRIZOL lysis followed by PureLink purification of the aqueous phase ( Life Technologies ) . 2 micrograms of total RNA was treated in a 10 microliter reaction with 0 units ( mock treatment ) or 20 units of RNase R ( Epicentre ) in 1× RNase R buffer , 1 unit/microliter murine Ribonuclease Inhibitor ( New England Biolabs ) , and incubated at 37DEGC for 1 hr . 1 microliter 1 mM EDTA , 1 microliter 10 mM each dNTP , and 1 microliter 100 microM random hexamer were added and the RNA denatured at 65DEGC for 5 min and placed on ice . 4 microliters 5× buffer ( 250 mM Tris-HCl pH 8 , 125 mM KCl , 15 mM MgCl_2 ) , 1 microliter murine Ribonuclease Inhibitor ( 40 units/microliter ) , and 1 microliter Superscript III ( LIfe Technologies ) were added; this cDNA reaction was incubated at 25 deg C 10 min , 50 deg C 50 min , 55 deg C 10 min , 85 deg C 5 min , 4 deg C hold . 0 . 5 microliter cDNA reaction was used as the template for qPCR and fraction resistant was computed as 2∧ ( RNase R C_t - Mock C_t ) . ” HeLa-S3 , A549 and AG04450 cells were grown in standard media and conditions . RNA was harvested by lysing cells with the TRIZOL reagent and purified on Purelink columns under ethanol concentrations that retain small and large RNAs . Total RNA reaction was reverse transcribed using the SuperScript III First-Strand Synthesis System ( Life Technologies , Carlsbad , CA ) with random hexamers according to the manufacturer's instructions . 500 ng/ul of cDNA was then used for each qPCR validation; dilution series were performed at concentrations of . 5 , 5 , 50 and 500 ng/ul . Each qPCR experiment was done in biological duplicate with 3 technical replicates each . For each cell type , we downloaded gtf files with gene level RPKM ( Reads per kilobase mapped ) estimates from the Poly ( A ) fraction; eg . for SK-NS-H_RA , we downloaded the file: http://hgdownload . cse . UCSC . edu/goldenPath/hg19/encodeDCC/wgEncodeCshlLongRnaSeq/wgEncodeCshlLongRnaSeqSknshraCellPapGeneGencV7 . gtf . gz RPKMs were summed across all genes to estimate total annotated poly ( A ) transcript abundance . In parallel , we summed all reads mapping to a circular RNA junction . These two values provide total abundance estimates of poly ( A ) and circular RNA respectively , up to a normalizing constant . We determined that normalizing constant by performing qPCR with two calibrating genes: FAT1 and HIPK3 . These genes were chosen for the reasons described in the main text . Standard curves were computed for each set of primers listed in Table S8 and used to compute relative expression of linear and circular RNA at the log scale . The difference between the log base 2 of calculated junctional circle counts and log base 2 RPKM and these differences was averaged for the 2 genes , and raised to the power 2 in order to normalize measurements . To test robustness of our estimates , we also performed the above analysis by imposing a filter on circles that could contribute to our estimate of total circle mass . That filter required a circular isoforms have greater than 5 counts in the cell type under consideration . This provided a conservative estimate of the moles of circular vs . poly ( A ) RNA . Using this filter , we obtained estimates of . 6% , 2% and . 6% for HeLa , A549 and AG04450 respectively , and is consistent with what we estimate without this filter . For each circular isoform represented by at least 50 counts in one sample ( and satisfying an FDR cut-off of . 025 ) , we fit a Poisson model per gene modeling circle counts by poly ( A ) gene expression ( genexp ) , celltype and total circles ( totcircles ) using the glm poisson model in R with the formula: cir∼log ( genexp ) +celltype+totcircles−1 . Coefficients in this model were used to choose genes shown in Figure 6 and the table of values is listed in Table S9 . 10 micrograms total RNA was denatured with glyoxal and run on a 2% agarose gel [36] , followed by alkaline capillary transfer onto Brightstar-Plus nylon membrane ( Ambion ) . 32P-labeled probe was generated by random-priming ( Prime-It II kit , Stratagene ) of a PCR product corresponding to exons 2 and 3 of CAMSAP1 ( nt 161–423 of GenBank # NM_015447 . 3 ) and hybridized in modified Church buffer ( 0 . 5M sodium phosphate pH 7 . 2 , 7% SDS , 10 mM EDTA ) at 65DEGC for 16 hr . After washing , the blot was visualized by phosphorimaging ( Typhoon , Molecular Devices ) . The following procedure was used to access statistical significance of the use of donor and acceptor sites across cell types . We analyzed donor and acceptor sites separately . Splice sites represented by more than 50 counts ( and satisfying an FDR cut-off of . 025 ) in at least one cell type were analyzed using this approach . For each such donor and acceptor site that was supported by more than 5 reads and present in at least two cell types , we computed an exact . 999 binomial confidence interval for its probability of use in that cell type . Sites with at least one pair of non-overlapping confidence intervals were identified and used to choose genes depicted in Figure 8 . Cell types were collapsed over replicates . A table of all confidence intervals by site and cell type is listed in Table S4 . Poly ( A ) depleted RNA isolated from Drosophila brain , available at ftp://ftp-trace . ncbi . nlm . nih . gov/sra/sra-instant/reads/ByStudy/litesra/SRP/SRP007/SRP007416/ was aligned to a custom database of annotated Drosophila exon-exon junctions using Jan . 2012 flyBase exon annotation and previously described methods and filters . A complete list of detected events is listed in Table S6 . To access evolutionary conservation , orthology of protein products was defined by Inparanoid using the following databases: http://inparanoid . sbc . su . se/download/7 . 0_current/sqltables/ For statistical assessment of expression of circular isoforms between mouse and human , a list of orthologous genes expressed ( measured by detected gene expression from RNA-Seq data sets used to measure circle abundance ) was compiled ( a total of 1402 genes ) . We then counted the number of genes in this table with more than 1 circle count in the mouse RNA-Seq data and with expression in the top 100 most expressed circular isoforms in one ENCODE ( human ) experimental replicate . 57 genes matched this criterion ( 4% ) . 147 genes on the list of 1402 were in the top 100 most expressed circular isoforms in one experimental replicate; 332 had more than 1 count in the mouse RNA-Seq data . Under an independence model , we expected 35 genes to match the joint criterion ( 2 . 5% ) . The residual from a chi-squared test for the independence model is ( O-E ) /sqrt ( E ) = 3 . 7 , which corresponds to a one sided p value of . 0001 . Read counts were summed across cell types and replicates for each isoform , defined here as a unique combination of gene , circularization splice donor coordinate , and circularization splice acceptor coordinate . Isoforms were filtered by requiring 20 or more read counts in total across ENCODE cell lines . Intron length was computed as described in the text . For each gene , intron lengths were considered either as fractions of the longest intron length within the gene , or as quantile ranks within the gene . Read counts were summed across replicates for each unique combination of cell type , gene , circularization splice donor coordinate , and circularization splice acceptor coordinate . Read counts were summed across cell types and replicates for each isoform , defined here as a unique combination of gene , circularization splice donor coordinate , and circularization splice acceptor coordinate . Genes where some annotated isoforms satisfied splice acceptor<donor and other isoforms satisfied donor<acceptor were removed from consideration; the remaining genes were then oriented such that all isoforms were acceptor upstream of donor . For each gene , the combinatorial index C compares the number of observed circular isoforms to the number of possible pairings of the detected acceptor and donor sites; C = 1 means that all possible pairings were actually detected , whereas C = 0 means that the minimum possible number of pairings was detected ( we adopted the convention that C is undefined when max . poss . isoforms = min . poss . isoforms ) . Precisely , C was defined for each gene as ( # of distinct circular isoforms detected – min . poss . isoforms ) / ( max . poss . isoforms – min . poss . isoforms ) , where min . poss . isoforms = max ( # of distinct acceptor sites detected , # of distinct donor sites detected ) , and max . poss . isoforms = # of combinations of 1 detected acceptor and 1 detected donor in which the acceptor is upstream of the donor . C is evaluated in Figure S4 . For each cell type and replicate , isoforms were defined here as unique combinations of gene , circularization splice donor coordinate , circularization splice acceptor coordinate , read 1 orientation , and read 2 orientation . Instances in which the junction-defining read ( “read 1” ) and its mate-pair read ( “read 2” ) were on the same strand were removed from consideration . For each cell type , replicates were pooled , and isoforms were defined here as unique combinations of gene , circularization splice donor , circularization splice acceptor , and read 1 orientation . The percentage of circular isoforms in the sense orientation ( with respect to the linear isoform ) is 96% for HMEC and >99% for the other 14 cell types . We downloaded a list of all high confidence microRNAs ( mature . fa from http://mirbase . org/ftp . shtml annotated as ‘Homo sapiens’ ) and corresponding 6mer seed match ( nt 2–7 ) . For each possible non-canonically ordered exon X , exon Y pair within a transcript in the UCSC knowngene transcript database ( enumeration beginning at 0 ) to 30 , we generated a corresponding 132 nt sequence consisting of 66 nt upstream and 66 nt downstream of the exon-exon junction . For each group of exon X -exon Y sequences , we compared the number of microRNA seed matches per offset ( from 0 to 126 ) divided by the total number of junctions evaluated . We compared these statistics for circular junctions expressed at rank <1000 in at least one cell type , ranked based on aligning paired end sequencing reads to a database of all UCSC knowngene exon-exon junctions and all other expressed circular junctions ( see Figure S5 ) . The rate of enrichment averaged 1 . 05 and was never more than 1 . 25 per offset . While this analysis does not strictly consider all microRNA binding sites within a circle , it samples a window including circular RNA sequence that , under most basic models where circular RNA were under selection to serve as a microRNA sink , would be enriched for microRNA seed matches . | Last year , we reported that circular RNA isoforms , previously thought to be very rare , are actually a pervasive feature of eukaryotic gene expression programs; indeed , the major RNA isoform from hundreds of human genes is a circle . Previous novel RNA species that initially appeared to be special cases , of dubious biological significance , have subsequently proved to have critical , conserved biological roles . An almost universal characteristic of regulatory macromolecules is that they are themselves regulated during development and differentiation . Here , we show that the repertoire of genes expressing circular RNA , the relative levels of circular: linear transcripts from each gene , and even the pattern of splice isoforms of circular RNAs from each gene were cell-type specific , including examples of striking regulation . In humans , we estimate that circular RNA may account for about 1% as many molecules as poly ( A ) RNA . The ubiquity of circular RNA and its specific regulation could significantly alter our perspective on post-transcriptional regulation and the roles that RNA can play in the cell . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| []
| 2013 | Cell-Type Specific Features of Circular RNA Expression |
Viral protein neutralizing antibodies have been developed but they are limited only to the targeted virus and are often susceptible to antigenic drift . Here , we present an alternative strategy for creating virus-resistant cells and animals by ectopic expression of a nucleic acid hydrolyzing catalytic 3D8 single chain variable fragment ( scFv ) , which has both DNase and RNase activities . HeLa cells ( SCH7072 ) expressing 3D8 scFv acquired significant resistance to DNA viruses . Virus challenging with Herpes simplex virus ( HSV ) in 3D8 scFv transgenic cells and fluorescence resonance energy transfer ( FRET ) assay based on direct DNA cleavage analysis revealed that the induced resistance in HeLa cells was acquired by the nucleic acid hydrolyzing catalytic activity of 3D8 scFv . In addition , pseudorabies virus ( PRV ) infection in WT C57BL/6 mice was lethal , whereas transgenic mice ( STG90 ) that expressed high levels of 3D8 scFv mRNA in liver , muscle , and brain showed a 56% survival rate 5 days after PRV intramuscular infection . The antiviral effects against DNA viruses conferred by 3D8 scFv expression in HeLa cells as well as an in vivo mouse system can be attributed to the nuclease activity that inhibits viral genome DNA replication in the nucleus and/or viral mRNA translation in the cytoplasm . Our results demonstrate that the nucleic-acid hydrolyzing activity of 3D8 scFv confers viral resistance to DNA viruses in vitro in HeLa cells and in an in vivo mouse system .
Viruses are pathogenic agents that cause potentially devastating diseases such as the flu , hepatitis , poliomyelitis , acquired immunodeficiency syndrome ( AIDS ) , severe acute respiratory syndrome ( SARS ) , avian influenza , and foot-and-mouse disease [1] , [2] , [3] . Many antiviral drug studies have been based on a functional analysis of viral genes and an understanding of the virus life cycle . McFarland and Hill ( 1987 ) showed successful vaccination of mice and pigs with a mutant PRV thymidine kinase [4] . Qing Ge also demonstrated that nucleocapsid siRNA or a component of the RNA transcriptase ( PA ) is a good antiviral drug to protect against influenza virus by inhibiting viral RNA transcription with siRNAs [5] . In addition , acyclovir , which is the best antiviral agent against HSV-1 , is a nucleotide analogue that shows an antiviral effect by inhibiting DNA replication [6] . However , commercially-developed antiviral drugs such as viral DNA polymerases , viral reverse transcriptases , and neuraminidase inhibitors target one or two viruses [7] , [8] , [9] , [10] , [11] . Thus , a new strategy is needed to prepare for outbreaks caused by new viruses or new mutant viruses because of the high mutation rates of viral genomes and recombination events among closely-related viruses [12] , [13] . A scFv is a recombinant antibody fragment , which commonly consists of a full variable region of an immunoglobulin heavy chain covalently linked to the corresponding variable region of an immunoglobulin light chain . scFvs have multiple benefits over traditional monoclonal antibodies due to their greatly reduced size , ease of genetic manipulation , and production of antibodies against viral proteins [14] . In 1994 , scFv which binds to the Human immunodeficiency virus 1 ( HIV-1 ) regulatory protein Rev was expressed intracellularly and potently inhibited HIV-1 replication in scFv immunized cells [15] . In addition , scFv against HIV integrase and reverse transcriptase showed reduced viral progeny in virus infected cells [16] , [17] . The retroviral capsid protein can be used as an antiviral target and thereby extend the number of targets that can potentially be used in combined scFv-based gene therapy approaches [18] . However , despite the many virus resistant studies using scFv proteins , no reports are available about scFv having an antiviral effect against a broad spectrum of viruses . Montandon et al . ( 1982 ) showed antiviral effects against Moloney murine leukemia virus ( M-MuLV ) with DNase I . In another study , DNase I digested DNA in the form of unmethylated proviral M-MuLV selectively [19] . Exonuclease , ISG20 , which is induced by type 1 interferon ( IFN ) , was over-expressed in CEL cells to inhibit HIV replication through nuclease activity [20] , [21] . Another case was reported in a plant system using Pac1 , an RNase isolated from the yeast Schizosaccharomyces pombe . Challenging a transgenic potato that expresses Pac1 with potato spindle tuber viroid ( PSTVd ) resulted in suppression of PSTVd infection and accumulation without significant undesired effects on the plant . The anti-viral effects of these RNases depend on their RNA-hydrolyzing activity [22] . We have shown previously that the 3D8 scFv protein has nuclease activity that non-specifically degrades in vitro DNA and RNA substrates [23] . 3D8 scFv originated from MRL mice and is a recombinant single chain antibody linked to VH and VL by linker peptides [14] . 3D8 scFv has catalytic activity supporting the hydrolysis of both single stranded and double stranded DNAs in the presence of Mg2+ without significant sequence specificity . In addition , efficient degradation of RNAs by 3D8 scFv in the presence of EDTA demonstrates that 3D8 scFv does not need divalent metal ions for its RNase activity [23] . Additionally , 3D8 scFv has DNase and RNase activity but when the active sites of each VL and VH are changed from histidine ( 35 position in VH and 94 position in VL ) to alanine , respectively , by site-directed mutagenesis , the mutant 3D8 scFv loses DNase activity but retains RNase activity [24] . We previously showed that 3D8 scFv can be applied to protect PK15 cells from virus infection . PK15 cells harboring 3D8 scFv show resistance to classical swine fever virus ( CSFV ) through RNase activity [25] . Also , we previously found that progenies of the transgenic tobacco plant acquired complete resistances against two single stranded ( ss ) -DNA geminiviruses , four ssRNA tobamoviruses , and one ssRNA cucumovirus [26] , [27] . So far , chemicals or specific antibodies for viral proteins have been mostly used in antiviral agents . But , 3D8 scFv with a completely different mechanism than antiviral agents is expected to be effective against a broad spectrum of viral infections . We established in vitro cell and in vivo mouse systems harboring 3D8 scFv genes and antiviral mechanism against PRV and HSV as DNA viruses with dsDNA genomes . The expression patterns of viral open reading frames ( ORFs ) were identified and the FRET assay and confocal microscopy were performed to investigate the antiviral mechanism shown in our experimental in vitro and in vivo systems . Taken together , our data support that the antiviral effect against the DNA virus used in this work was caused by ( 1 ) nuclear DNase activity that inhibited DNA replication or RNA transcription and ( 2 ) RNase activity in the cytoplasm blocked protein translation .
The gene encoding 3D8 scFv was introduced into HeLa cells to test whether 3D8 scFv expressing cell lines could protect against DNA virus ( HSV-1 and PRV ) infection . The DNAs coding for both the wild-type 3D8 scFv protein ( SCH ) and the inactive DNase-mutated 3D8 scFv protein ( muSCH ) were cloned under the transcriptional dependence of the cytomegalovirus promoter in the pcDNA3 . 1/V5-HisB vector ( Figure 1A ) . Three 3D8 scFv lines ( SCH07041 , SCH07071 , and SCH07072 ) and mutated 3D8 scFv ( muSCH ) were selected by serial dilution in media containing G418 [28] . We evaluated 3D8 scFv expression levels in selected cell lines by quantitative real-time reverse transcription polymerase chain reaction ( RT-PCR ) , flow cytometric analysis , and immunocytochemistry . According to quantitative real-time RT-PCR , SCH07072 expressed the highest 3D8 scFv RNA compared to that of the other two cell lines ( 1 . 3 and 4 times more ) . The mutant 3D8 scFv-transfected ( muSCH ) cells expressed similar levels of 3D8 scFv RNA to SCH07072 ( Figure 1B ) . Flow cytometry data revealed that when the mean for the HeLa cell line as a control was 79 . 48 , SCH07041 , SCH07071 , and SCH07072 were 88 . 45 , 109 . 17 , and 126 . 38 , respectively ( Geo mean: HeLa = 74 . 51 , SCH07041 = 76 . 88 , SCH07071 = 89 . 89 , and SCH07072 = 115 . 33 ) and was similar to the results shown in the quantitative real-time RT-PCR analysis ( Figure 1C ) . Confocal microscopy demonstrated that 3D8 scFv proteins in SCH07072 and muSCH were targeted and localized in both the cytosol and nucleus ( Figure 1D ) . Common features of anti-DNA antibodies that have a large number of positively-charged residues within complementarity determining regions ( CDRs ) of VH or VL domains due to their antigen binding properties , which resemble the nuclear localization signal ( NLS ) , have been attributed to their final accumulation within the nucleus of cells , such as cell-penetrating peptides ( CPPs ) [29] . Therefore , 3D8 scFv proteins were localized in the cytosol using a vector system and targeted to the nucleus by a NLS . The three 3D8 scFv lines and the one mutant 3D8 scFv line were challenged with HSV-1::GFP and PRV which had dsDNA as a genome . Virus-infected HeLa cells express GFP in the cytoplasm after HSV infection [30] and PRV infection results in cytopathic effects ( CPE ) in HeLa cells . In addition , PRV results in multinuclear giant cell formation as a PRV CPE [31] . GFP expression levels in the four cell lines challenged with three different MOI ( 0 . 1 , 0 . 5 , and 1 . 0 ) of HSV::GFP were observed under a fluorescence microscope 48 hours after virus challenge . The SCH07072 line showed the lowest GFP expression compared to that of the other two lines ( SCH07041 and SCH07071 ) and the mutant line ( muSCH ) ( Figure 2A ) . The plaque assay revealed that SCH07072 cells produced the fewest number of viral progeny of 163/ml at an MOI of 0 . 1 and 530/ml at an MOI of 0 . 5 . The plaque assay data supported the GFP expression results showing the best antiviral effects among the three cell lines . In contrast to SCH07072 , wild-type HeLa cells had 1 , 343/ml ( 8 . 24 times higher ) and 7 , 851/ml ( 14 . 81 times higher ) viral progeny at MOIs of 0 . 1 and 0 . 5 , respectively ( Figure 2B ) . At an MOI of 0 . 5 , muSCH cells produced a viral titer that was 14 . 42 times higher than that of SCH07072 and was similar to that of wild type HeLa cells ( Figure 2B ) . Western blot analysis using an anti-HSV DNA polymerase antibody ( POL/UL42 complex ) demonstrated that less HSV-1 DNA polymerase was detected in SCH07072 cells compared to that of wild-type HeLa cell lines 48 hours after virus challenge ( Figure 2C ) . Similar to HSV , PRV also showed similar antiviral effects in the three 3D8 scFv cell lines and one mutant cell line . Forty-eight hours after the PRV challenge , fewer cytopathic effects , as assessed by multinuclear giant cell formation , were observed in SCH07072 cells compared to the other types of cells , including wild-type cells ( Figure 3A ) . The plaque assay also showed the highest antiviral effects in SCH07072 cells with viral titers of 916/ml and 7 , 466/ml at MOIs of 0 . 1 and 0 . 5 , respectively , compared to those of wild-type HeLa cells ( 9 , 130/ml and 31 , 833/ml , respectively ) and the muSCH cell line ( 9 . 046/ml and 30 , 400/ml , respectively ( Figure 3B ) . The expression levels of 3D8 scFv in SCH07071 and SCH07072 cells were similar ( Figure 1B ) but the antiviral activity of SCH07072 and SCH07071 were quite different . Generally there is a correlation between 3D8 scFv expression and antiviral activity . However , in order for cells to acquire the antiviral activity of 3D8 scFv , a certain amount of 3D8 scFv protein needs to be expressed . We thought that there might be a threshold for the acquired antiviral activity conferred by 3D8 scFv expression . SCH07072 may have reached the threshold but SCH07071 did not ( Figure 3B ) . Glycoprotein D ( gpD ) was used for Western analysis to count the PRV . The gpD protein was not found in the SCH07072 lines but wild-type HeLa cells showed strong signals ( Figure 3C ) . Even though the expression level of 3D8 scFv in SCH07041 cells was significantly lower ( approx . 1/3 ) than that of SCH07072 cells , both cells were found to be equally resistant against HSV-1 infection at an MOI of 0 . 1 ( Figure 2B ) , suggesting that the 3D8 scFv protein level in SCH07041 cells was close to the maximum effective dose against HSV-1 . At a higher virus titer ( MOI 0 . 5 ) , the antiviral effect of 3D8 scFv was shown in a more dose-dependent manner such that much less viral infectivity was found in the SCH07072 cells than the SCH07041 ( 1 , 964 . 00 vs 530 . 67 pfu/ml ) . This was further demonstrated in case of PRV infection study ( Figure 3B ) , in which the SCH07072 cells were shown to be more resistant against the virus infection than the SCH07041 cells at both MOI of 0 . 1 and 0 . 5 . As a result , 3D8 scFv expression levels and antiviral effects showed a directly proportional relationship . The FRET assay was carried out to confirm that 3D8 scFv nuclease activity was responsible for the antiviral activity shown by SCH07072 cells [32] ( Figure 4 ) . SCH07072 cells produced up to 660 relative fluorescence units ( RFU ) 80 min after the 6-carboxyfluorescein ( FAM ) -labeled DNA substrate was introduced into the cells . However , the fluorescence output from HeLa cells , muSCH cells , and the negative control reactant was a maximum of 200 RFU or less ( Figure 4A ) . When the FAM-labeled RNA substrate was introduced into cells , SCH07072 and muSCH cells had fluorescence levels of about 200 RFU , but HeLa cells and the negative control reactant showed fluorescence levels of only 80 RFU ( Figure 4B ) . These FRET results confirmed that 3D8 scFv in SCH07072 had both DNase and RNase activity , but that only RNase activity was observed in muSCH cells ( Figure 4 ) . Therefore , the protection against the virus infection shown by SCH07072 cells was due to 3D8 scFv nuclease activity against the virus genome itself . Quantitative RT-PCR was performed to investigate how 3D8 scFv protected against virus infection in relation to the virus infection cycle . The amounts of viral DNA and RNA isolated from cells grown with or without HSV DNA polymerase inhibitor ( Phosphonoacetic acid , PAA ) in the culture media were analyzed by quantitative real-time PCR ( Figure 5 ) . PAA inhibited the synthesis of HSV DNA in infected cells and the activity of the virus-specific DNA polymerase in vitro [33] . PAA was used to investigate the antiviral effects by 3D8 scFv DNase activity in the immediate early stage before de novo viral DNA replication . The changes in viral DNA content in HeLa , SCH07072 , and muSCH cells were traced at immediate early ( 2 hr after virus challenge , ICP4 ) , early ( 6 . 5 hr after virus challenge , UL9 ) , and late ( 25 hr after virus challenge , UL19 ) stages of infection ( Figure 5A ) . Viral DNA accumulation in SCH07072 cells was reduced by up to 94 . 2% under the PAA-untreated condition , compared to HeLa cells at each stage . Virus DNA accumulation in SCH07072 cells cultured with PAA decreased by up to 80% compared to that in HeLa cells at all stages of infection . In contrast , muSCH cells under the PAA-untreated condition showed as much virus DNA accumulation as HeLa cells at the early stage of infection , but 68% more viral DNA was detected in muSCH cells compared to HeLa cells at the late stage . The accumulation of virus DNA in SCH07072 indicates that nuclear 3D8 scFv can degrade viral DNA directly , resulting in less accumulation of viral DNA ( Figure 5B ) . Viral RNA expression was also analyzed under the same experimental conditions . Viral RNA accumulation of six ORFs decreased rapidly in PAA-treated cells at both the early and late stages of infection ( Figure 5C ) . Only SCH07072 cells showed a substantial reduction in viral RNA accumulation during the entire infection cycle under the PAA-untreated condition compared to HeLa and muSCH cells . The expression of ICP0 and ICP4 in SCH07072 cells cultured with PAA decreased by 72 . 5% and 74 . 9% , respectively , compared to HeLa cells during the immediate early stage . At the late infection stage , SCH07072 cells expressed 83 . 1% and 87% less UL19 and UL 38 transcripts , respectively , than those of HeLa cells after virus challenge ( Figure 5C ) . However , ICP0 and ICP4 levels were 69 . 4% and 60 . 7% lower than in HeLa cells during the immediate early stage of muSCH cell infection , and were similar to the viral RNA content seen in SCH07072 cells . However , at the late stage of infection , the viral RNA accumulation patterns in muSCH cells were similar to those in HeLa cells ( Figure 5C ) . The HSV-1 RNA transcription occurs during immediate early stage of the viral infection stage together with a limited level of its DNA replication , as summarized on Edward K . Wagner's web site ( http://darwin . bio . uci . edu/~faculty/wagner/hsv4f . html ) . Therefore , in the absence of PAA the decreased level of viral DNA in muSCH cells in comparison with that of HeLa cells is likely to be attributed to an interference of HSV -1 transcription by the RNase activity of the mutant 3D8 scFv ( Figure 5B ) . With treatment of PAA , a specific inhibitor of viral DNA polymerase , the level of viral DNA replication in muSCH and HeLa cells were shown to be almost identical at immediate early stage . This relatively mild impact of PAA on DNA replication in muSCH cells as compared with HeLa and SCH07072 cells was also observed when selected viral RNA transcripts were measured in these cells at immediate early stage ( Figure 5C ) . Transcriptions of these genes were found to be significantly decreased in HeLa and SCH07072 cells upon PAA , which reflects the decreased level of viral DNA templates for transcription . By contrast , transcription level of these genes in muSCH cells at the same stage did not show any significant difference between PAA treated and untreated cells . On entry into the nucleus , the genome of HSV-1 is also associated rapidly with histone proteins [34] and nucleosomes are assembled , though in irregularly or randomly spaced manners , on 1 hour post infection . Therefore , it is intriguing and difficult to explain on why and how only the viral genome is subjected to the nucleic acid hydrolyzing activity of 3D8 scFv , leaving host genome largely unaffected . Perhaps as the HSV-1 genome , which was not completely assembled with nucleosomes or was assembled with unstable nucleosomes , can be digested by intrinsic DNase of 3D8 scFv during HSV-1 replication takes place ( 6 hours post infection or early stage ) , especially considering that the viral genome is found to be relatively free of histones following viral DNA replication , beyond 6 hours post infection [35] . When we compared the HSV-1 DNA accumulation between SCH07072 and muSCH cells on immediate early stage , early stage and late stage respectively , muSCH cells showed limited antiviral activities at immediate early stage but did not show any antiviral activity later on . Thus it may indicates that mutant 3D8 scFv with only intrinsic RNase activity is not sufficient for providing a full protection for host cells against HSV-1 infection once HSV-1 DNA replication successfully takes places in nucleus in muSCH cells . Therefore , it may be possible that the antiviral effects of 3D8 scFv against HSV-1 and PRV infection were contributed by both the DNase activity , which is mainly effective in the nucleus and the RNase activity mainly effective in the cytoplasm . This proposed antiviral activity can be explained by ( 1 ) temporal blockage of viral DNA replication in the nucleus and protein translation in the cytosol and ( 2 ) spatial protection ( nucleus vs . cytosol ) provided by 3D8 scFv . However , further experimental investigations are needed to confirm the antiviral mechanisms of 3D8 scFv proposed in this study . The 3D8 scFv protein shown in SCH07072 was localized in both the cytoplasm and nucleus ( Figure 1D ) . To test our hypothesis that 3D8 scFv acted as a DNase in the nucleus , we investigated whether 3D8 scFv could hydrolyze histone-bound DNA and methylated DNA . Both 3D8 scFv and DNase I at 10×10−4 U/µl started to hydrolyze 0 . 2 µg DNA 1 hr after treatment . At a concentration of 8 . 3×10−4 U/µl , 3D8 scFv digested more non-methylated DNA than methylated DNA after 1 hr of treatment ( Figure 6A ) . In addition , quantitation of non-methylated and methylated DNA showed a reduction rate of 62% and 21% , respectively ( Figure 6C ) . However , no difference in hydrolysis of non-methylated and methylated DNA was observed when 8 . 3×10−4 U/µl DNase I was tested ( Figure 6A ) . Both 3D8 scFv and DNase I did not digest DNA at a concentration of 1 . 4×10−4 U/µl , regardless of whether it was methylated or not ( Figure 6A ) . We next tested whether the 3D8 scFv protein could digest histone-bound DNA in the nucleus . Histone-bound DNA fragments ( 150 bp ) were not hydrolyzed by either 3D8 scFv or DNase I at concentrations of 10×10−4 U/µl and 8 . 3×10−4 U/µl , as shown in Figure 6B . But , at concentrations of 8 . 3×10−4 U/µl 3D8 scFv , DNA without histones was reduced 34% and 90% after 1 hr and 3 hr , respectively ( Figure 6C ) . Also , 3D8 scFv digested more DNA without histones than DNase I at concentrations of 8 . 3×10−4 U/µl 1 hr after treatment ( Figure 6B ) . The nuclear DNA of eukaryotic cells is mostly modified by methylation and interacts with various nuclear proteins such as histone proteins [36] , [37] , [38] , [39] . Figure 6 shows that 3D8 scFv did not hydrolyze methylated or histone-bound DNA when the activity of 3D8 scFv was <1 . 4×10−4 U/µl . This result indicates that chromosomal DNA in the nucleus was protected from 3D8 scFv because chromosomal DNA is mostly methylated and histone-bound . , 3D8 scFv proteins in SCH07072 cells are expressed at lower working concentrations than we tested in our in vitro analysis shown in Figure 6 . HeLa , SCH07072 , and muSCH cells showed similar doubling patterns when 72 hr growth curves were plotted ( Figure S1A ) . Northern blot analysis showed that the RNA stability and RNA expression patterns of three marker genes ( GAPDH , actin , and VEGF ) in the three cell lines were not affected by expression of the 3D8 scFv protein ( Figure S1B ) . Taken together , expression of the 3D8 scFv protein in SCH07072 cells did not affect cell growth or viability . Pac1 and ISG20 , which have RNase activity , have been used as antiviral proteins against potato spindle tuber viroid and HIV , respectively [21] , [22] . Our observations and those of others on the effects of exogenous expression of RNase ( Pac1 and ISG20 ) and/or nuclease ( 3D8 scFv ) indicate that it is possible to develop transgenic plants and animal cells and even entire organisms that are virus-resistant . However , host RNA could be degraded non-selectively by 3D8 scFv RNase activity in 3D8 scFv-expressing cells . During the transformation process , antiviral transgenic cell lines or organisms can be developed if ( 1 ) the amount of 3D8 scFv protein expressed does not inhibit the physiological and developmental processes of cells or organisms , and ( 2 ) the amount of 3D8 scFv protein produced is sufficient to protect against virus infection . 3D8 scFv and DNase I were transferred to HeLa cells using a microporator to investigate cell viability using a neutral red assay . The concentration of each protein was adjusted to be 1 . 5625×10−3 to 0 . 05 units . No cell viability differences were observed between 3D8 scFv and DNase I for 3D8 scFv and DNase I concentrations of up to 0 . 05 units ( Figure S2A ) . 3D8 scFv and DNase I were detected in both the cytosol and nucleus of HeLa cells under confocal microscopy ( Figure S2B ) . Montandon et al . ( 1982 ) reported the antiviral effects of DNase I against M-MuLV in mouse cells . DNase I digested non-methylated DNA of proviral M-MuLV but did not hydrolyze methylated endogenous M-MuLV [19] . However , when we introduced the same unit amounts of DNase I and 3D8 scFv into HeLa cells using a microporator , the antiviral effects of the 3D8 scFv protein were almost 8-fold higher than those of DNase I ( Figure S2C ) . This result indicates that 3D8 scFv has a greater antiviral effect than DNase I , most likely due to the additional RNase activity of 3D8 scFv ( Figure S2 ) . After the antiviral effects of the 3D8 scFv protein was confirmed in HeLa cells , transformed mice harboring the 3D8 scFv gene were produced to investigate the antiviral effects in an in vivo mouse system . The pcDNA3 . 1-3D8 scFv plasmid was linearized with NruI/StuI/PvuI enzymes and then used for transformation of C57BL/6NCrjBgi . A total of 150 F0 were generated and 10 lines ( F0: 47 , 69 , 90 , 92 , 108 , 109 , 110 , 115 , 128 , and 135 ) out of 150 F0 were selected by PCR analysis ( Figure S3A ) . Each transformant was reconfirmed by Southern hybridization after genomic DNA was digested by EcoRI and HindIII enzymes . All transformants showed single bands that reacted with the probes prepared with the 3D8 scFv full gene ( Figure S3B ) . 3D8 scFv expression levels were analyzed in the different transgenic lines and different organs . The muscle , brain , and liver were chosen because these organs are virus infection routes and the relative 3D8 scFv gene expression levels were investigated by quantitative real-time RT-PCR ( Figure 7A ) . The F0 69 , 90 , and 135 lines were selected from 10 F0 lines for further analysis based on 3D8 scFv expression levels in the TG mice lines and previous therapeutics experiments , which showed that PRV was detected in muscle and brains after mice were challenged with PRV and 3D8 scFv proteins by intramuscular and intraperitoneal injections , respectively . Therefore , the F0 90 line , which expressed the highest 3D8 scFv levels in both brain and muscle , and the F0 135 line , which expressed high amounts of 3D8 scFv in the brain , were selected and used for further investigations on the antiviral preventive effects . The F0 69 line showed positive gPCR and RT-PCR results but 3D8 scFv expression levels in the F0 69 line were low; thus , it was used as a control . On the basis of the expression levels in the F0 69 line , the F0 90 line produced 12 . 25 times more 3D8 scFv in the liver and the F0 110 and F0 135 lines showed 0 . 29 and 0 . 07 times less than that of the F0 69 line . However , the F0 90 and F0 135 lines had 3 , 270 and 3 , 174 times more 3D8 scFv in the brain than that of the F0 69 line , even 201 . 98 times more in F0 110 . Therefore , the F0 69 line was used as a control and the F0 90 and F0 135 lines were used for testing the antiviral preventive effects in TG mice lines after PRV infection ( Figure 7A ) . The 3D8 scFv TG founder ( F0 lines ) was mated with wild-type C57BL/6NCrjBgi to establish the lines and produce siblings from the F0 lines . Four F1 lines ( 69 F1: 31 , 33 , 38 , 208 ) were produced from the F0 69 lines . Each of the four F169 lines were mated with wild-type C57BL/6NCrjBgi and then 20 F269 lines were produced . They were used for virus challenge experiments and named STG69 . In the case of the 90 and 135 F1 lines , three F190 lines ( F190: 45 , 46 , and 162 ) and five F1135 lines ( F1135: 140 , 142 , 144 , 165 , and 166 ) were produced and then 17 F290 lines ( STG90 ) and 24 F2135 lines ( STG135 ) were prepared ( Figure S4A ) . Genomic PCR was performed to confirm the transgenic lines at each generation from F0 to F2 ( data not shown ) . The number of live and dead mice was counted every 12 hr for 5 days to investigate the viability in each TG line after the femoral muscle was challenged with 10 LD50 PRV was ( Figure S4B ) . Figure 7B shows that the viability of WT mice challenged with PRV was 11% ( 2/18 ) . The viabilities of the STG69 and STG135 lines against PRV challenges were about 40% ( 8/20 ) and 17% ( 4/24 ) , respectively . However , the STG90 line showed 53% ( 9/17 ) viability . These data were described by Kaplan–Meier analysis . Mice generally began to exhibit clinical signs of illness 3–5 days post-challenge ( Figure 7C ) . RT-PCR and immunohistochemistry were performed to confirm virus accumulation in the three 3D8 scFv TG lines after PRV inoculation . Figure 8A shows that PRV gpD RNA was detected in the muscle and brain of WT , STG69 , and STG135 mice accompanied by disease symptoms . Immunohistochemical analysis with an anti-gpD antibody supported the viral gene expression data; only WT mice with a high infection rate had Purkinje layer cells from the brain that stained dark brown after diaminobenzidine ( DAB ) staining ( Figure 8B ) . These data indicate that virus multiplication was not inhibited by the 3D8 scFv protein if virus infection was not protected against in the early infection stages by the 3D8 scFv protein . This also indicates that PRV DNA was hydrolyzed immediately in virus-inoculated STG90 muscle cells , and that the virus titer was not high enough to infect muscle cells . Therefore , PRV could not move systemically into the spinal cord or brain in STG90 mice . However , STG135 mice were not protected against PRV infection even though 3D8 scFv protein was expressed in the brain ( Figure 7A ) at similar levels to those observed in STG90 mice . We interpret these data to indicate that the expression and presence of antiviral proteins in inoculated cells and tissues is as important as high expression levels in target cells or tissues . Blood biochemistry ( Table 1 ) and body weight changes for 7 weeks after birth ( Figure S5A and S5B ) were not significantly different between STG90 and WT mice . Also , we identified the expression patterns of endogenous genes including apoptosis ( K-ras and BAX ) , growth factor ( VEGF ) , and housekeeping ( GAPDH ) genes between STG90 and wild-type mice; however , there was no difference ( Figure S5C ) . Although we have demonstrated the antiviral effect attributed by DNase and RNase activity of 3D8 scFv without any cytotoxic effect on the host cells in this study ( Figure S1A and S1B ) , possibility of the host DNAs damaged by non-specific nuclease activity of 3D8 scFv cannot be entirely ruled out . In our previous study , when we tested the potential cytotoxic effect of 3D8 scFv on HeLa cells with varying concentration of 3D8 scFv ( 5–40 µM ) for 48 hrs , the cell viability was found to drop by ∼50–60% upon 10 µM of 3D8 scFv treatment , while treatment of the cells with 1 µM and 5 µM 3D8 scFv exhibited no significant cytotoxicity up to 48 h of incubation ( 0 and ∼20% respectively ) [40] . This may indicate that the host cells indeed begin to be affected by the non-specific activity of 3D8 scFv at higher dosages ( i . e . , 5 µM and up ) . Although it is not clear on exactly how much 3D8 scFv protein actually penetrated into the cells from media when 5 µM 3D8 scFv protein was administered , 5 µM 3D8 scFv is fairly large amount of proteins for in vitro cytotoxicity analysis even though we used the same amount of the proteins for in vivo survival experiments with mice ( Figure 7 ) . On the other hand , the actual level of 3D8 scFv expressed in transgenic cells such as in SCH07072 cell line was too low to be even detected by Western blot analysis , although its expression was confirmed by confocal microscopic observation and flow cytometry analysis ( Figure 1 ) . Therefore , it can be concluded that the low concentration of 3D8 scFv expressed in transgenic cell lines or TG mice is generally sufficient for conferring antiviral effects without incurring damage to the host DNA , but the likelihood of 3D8 scFv having deleterious impact on host cells as well through non-specific nucleic acid hydrolyzing activity remains valid , especially at higher concentration of the protein . In conclusion , the DNA virus-protective effects conferred by the 3D8 scFv protein can be attributed to the DNase activity of the protein in the nucleus and RNase activity in the cytoplasm . This antibody inhibits ( 1 ) viral DNA replication and RNA transcription in the nucleus by viral DNA degradation and ( 2 ) viral protein translation in the cytoplasm via viral RNA degradation . In other words , the 3D8 scFv protein attacks the virus DNA genome itself or its RNA transcripts in two different subcellular spaces ( nucleus and cytoplasm ) and at two different times ( viral replication and viral transcription ) ( Figure 9 ) . These antiviral effects of 3D8 scFv have not only been observed against viruses with a dsDNA genome but also ssRNA viruses such as CSFV [25] and several plant viruses such as geminivirus , tobamovirus , and cucumovirus [26] , [27] . Taken together , 3D8 scFv is a candidate antiviral protein that can potentially confer resistance to a broad spectrum of animal and plant viruses .
HeLa cells were provided by the Korean Cell Line Bank and were maintained in DMEM medium supplemented with 10% fetal bovine serum ( Hyclone , Logan , UT , USA ) , 100 U/ml penicillin-streptomycin ( Hyclone ) , and non-essential amino acids ( Sigma , St . Louis , MO , USA ) , at 37°C in a 5% CO2 atmosphere . HSV-GFP virus was obtained from the American Type Culture Collection ( ATCC Number: VR-1544 ) . The PRV-YS strain was obtained from the National Veterinary Research and Quarantine Service ( NVRQS ) of Korea . Anti-3D8 scFv polyclonal antibody was provided by Dr . Kwon ( Ajou University School of Medicine ) . Anti-PRV gD monoclonal antibody was purchased from Jeno Biotech Inc . ( Chuncheon , Korea ) and anti-HSV monoclonal antibody was obtained from Chemicon ( Temecula , CA , USA ) . Anti-GAPDH polyclonal antibody was purchased from Santa Cruz Biotechnology ( Santa Cruz , CA , USA ) . 3D8 scFv and mu3D8 scFv were amplified by PCR from the pIg20-3D8 scFv vector [23] and then subcloned into the pcDNA3 . 1 V5/His-B vector ( Invitrogen , Carlsbad , CA , USA ) . pcDNA3 . 1-3D8 scFv and pcDNA3 . 1-mu3D8 scFv were transfected into HeLa cells using Fugene HD transfection reagents ( Roche , Indianapolis , IN , USA ) and selected by G418 selection ( 400 µg/ml ) . The relative concentrations of 3D8 scFv and mu3D8 scFv were calculated by quantitative real-time PCR after normalization to the GAPDH gene ( Table 2 . ) . The primer efficiency of 3D8 scFv and GAPDH are 1 . 845 and 2 . 065 , respectively . Confocal microscopy and flow cytometry were performed as described previously [40] . Cells on coverslips were washed in phosphate-buffered saline ( PBS ) and fixed for 10 min in 4% paraformaldehyde in PBS at room temperature . Cells were permeabilized with Perm-buffer ( 1% BSA , 0 . 1% saponin , 0 . 1% sodium azide in PBS ) for 10 min at room temperature ( RT ) . After blocking with 3% BSA in PBS for 1 hr , 3D8 scFv-treated cells were incubated with rabbit anti-3D8 scFv Ab , followed by TRITC-anti-rabbit Ig . Nuclei were stained with DAPI during the last 10 min of incubation at RT . Cells on coverslips were mounted in Vectashield anti-fade mounting medium ( Vector Labs , Burlingame , CA , USA ) , and observed with a Zeiss LSM 510 laser confocal microscope and analyzed with Carl Zeiss LSM Image software ( Jena , Germany ) . 3D8 scFv-transfected cell lines were assessed by flow cytometry . HeLa cells grown in six-well plates ( 1×105 cells/well ) were pre-incubated in serum-free DMEM for 30 min at 37°C and were either untreated or treated with each inhibitor for 30 min at 37°C before 3D8 scFv ( 10 mM ) treatment . Cells suspended with trypsin were treated once more with 0 . 1% trypsin for 3 min at 37°C to wash off surface-bound proteins . After washing with ice-cold PBS once , cells were fixed and permeabilized as per the procedures described above . Cells were washed with ice-cold PBS twice , labeled with anti-rabbit 3D8 scFv Ab , followed by TRITC-anti-rabbit Ab , and then analyzed using a FACS Calibur TM ( BD Biosciences , San Diego , CA , USA ) . A total of 1×104 cells were analyzed for each test . Quantification of HSV-1 DNA and RNA from cells of SCH07072 , muSCH , and HeLa was carried out by harvesting the cells at 2 hours post infection ( HPI ) with HSV-1 , 6 . 5 HPI , and 25 HPI for immediate early stage , early stage , and late stage , respectively as summarized in Figure 5A . RNA was isolated from cells and mouse tissues ( liver , muscle , lung , and brain ) using Corezol reagent ( CoreBio System ) [41] . cDNA was synthesized from 5 µg of total RNA using random hexamers and MMLV reverse transcriptase ( SuperBio ) . All primers were designed using the Primer 3 program . The expression levels of immediate early genes [ICP0 ( transcriptional transactivator ) and ICP4 ( regulatory protein ) ] [42] , [43] , early genes [UL9 ( replication origin binding ) and UL29 ( single-strand binding protein ) ] [44] , [45] and late genes [UL19 ( capsid protein ) and UL38 ( capsid assembly ) ] [46] , [47] were analyzed by Rotor-Gene 3000 ( Corbett Research , Sydney , Australia ) ( Table 2 ) . The primer efficiency of ICP0 , ICP4 , UL9 , UL29 , UL19 , and UL38 are 2 . 081 , 1 . 989 , 1 . 972 , 1 . 916 , 1 . 983 , and 2 . 053 , respectively . Each data point represents the average of three individual experiments and the error bars indicate standard errors . Virus-infected HeLa cells and SCH07072 cells were lysed using Pro-prep ( Intron , Daejeon , Korea ) . Cell lysates were analyzed by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis , transferred to a nitrocellulose membrane using a semi-dry method , incubated with primary antibody ( 1∶500 dilution for cleaved caspase 3 and INF-beta and a 1∶1000 dilution for PRV gD and HSV ) for 12 hr at 4°C , and then incubated with goat anti-mouse or rabbit IgG-HRP conjugate ( 1∶1000 ) for 1 hr at 37°C . A 21 base-long ribo-oligonucleotide ( 500 nM ) labeled with 6-carboxyfluorescein ( FAM ) at the 5′ terminus and a black hole quencher ( BHQ ) at the 3′ terminus were synthesized ( 5′ FAM-CGATGAGTGCCATGGATATAC-BHQ 3′ ) , annealed to the non-labeled complementary ribo-oligonucleotide , and then delivered into cells in 96-black-well plates using Lipofectamine transfection reagent ( Invitrogen ) . Immediately after changing the medium , the real-time fluorescence intensity of the cells was read for 2 hr in real time using a fluorescence analyzer at 5 min intervals ( Molecular Devices , Sunnyvale , CA , USA ) . Vero cells were cultured on 6-well plates ( Nunc ) . Ten-fold serial dilution of the virus samples was prepared in serum-free DMEM media . Each dilution ( 100 µl ) was used to infect Vero cells in duplicate . The virus was allowed to adsorb at 37°C for 1 hr . After 48 hr incubation , the Vero cells were rinsed twice with PBS and stained with crystal violet solution for 5 min . The plates were washed with ddH2O and the virus titer was calculated . This plaque analysis included four biological and three technical replicates . Each cell line was infected by virus four times and the numbers of plaques on Vero cells were counted three times , respectively , by using the culture media containing viruses . HeLa genomic methylated DNA and non-methylated DNA were purchased from NEB . Each DNA type ( 0 . 2 µg ) was treated with 3D8 scFv and DNase I at three different unit concentrations ( 10×10−4 , 8 . 3×10−4 , and 1 . 4×10−4 U/µl ) . HeLa chromatin was prepared using the EZ-Zyme Enzymatic Chromatin Prep kit ( Upstate Technology , USA ) [48] . Prepared chromatin DNA and naked DNA were treated with 3D8 scFv and DNase I ( 10×10−4 U/µl and 8 . 3×10−4 U/µl , respectively ) . The DNA samples were harvested after 0 , 1 , 2 , and 3 hr of incubation and then each DNA sample was analyzed on an agarose gel for abzyme analysis . The relative quantification of DNA was measured by the height values and performed using GeneSnap 7 . 09 software ( SynGene , UK ) . A neutral red ( NR ) cytotoxicity assay was performed to test cell viability after treatment with DNase I or 3D8 scFv . DNase I and 3D8 scFv at concentrations ranging from 1 . 5625×10−3 units to 0 . 05 units were transferred to HeLa cells using a microporator ( iNCYTO ) . Uptake of neutral red dye into intracellular acidic compartments was determined by measuring absorbance at 540 nm . 3D8 scFv transgenic mice were produced by Macrogen Co . using standard microinjection procedures . Briefly , fertilized mouse eggs were flushed from the oviducts of superovulated C57BL/6NCrjBgi mice , and male pronuclei were injected with a 2 . 6 kb fragment ( 4 ng/µl ) of 3D8 scFv that had been obtained by digesting the pcDNA3 . 1/V5-His B ( 3D8 scFv ) vector with NruI/StuI/PvuI restriction enzymes . The injected eggs were reimplanted in the oviducts of pseudo-pregnant C57BL/6NCrjBgi recipient females . At 3 weeks of age , the animals were tested for the presence of the transgene by PCR analysis of their genomic DNA using forward ( 5′ CAGAGCTCTCTGGCTAACTAG 3′ ) and reverse primers ( 5′ CTGTTGAACAGACTCTGACTG 3′ ) . The 3D8scFv TG founders ( F0 lines ) were mated with wild-type C57BL/6NCrjBgi mice to establish the transgenic lines and produce siblings from the F0 lines in a breeding program provided by Macrogen Co . The principles of laboratory animal care ( NIH publication 85-23 , revised 1986 ) were followed and all experiments were carried out under the guidelines of the NVRQS , Korea . Southern blot hybridization was used to confirm integration and determine the copy number of the 3D8 scFv gene in the transgenic mice . Genomic DNA ( 20 µg ) from each transformant was digested with EcoRI and HindIII restriction enzymes and then the DNA was separated on 1% agarose gels . 3D8 scFv-specific [32P]-radiolabeled probes were prepared from the pcDNA3 . 1V5/HisB-3D8 scFv vector . Three 3D8 scFv TG mouse lines and the wild-type C57BL/6 line were divided into five groups with or without PRV infection ( WT-mock , WT-PRV , STG69-PRV , STG90-PRV , and STG135-PRV ) . Mice were challenged with PRV by intramuscular injection ( 10 LD50 ) . Survival rates in each group were calculated at 12 hr intervals ( Figure S5 ) . Tissues ( muscle and brain ) were harvested and stored at −70°C for further RT-PCR and immunohistochemical analyses . Representative 3 µm-thick brain tissue sections for immunohistochemical analysis were mounted on silane-coated slides as described by Ramos-Vara [49] . Anti-PRV gpD antibody ( Jeno Biotech Inc . ) was used as the primary antibody ( 1∶100 ) and was applied for 60 min at RT . The samples were then treated for color development with the DAB Detection Kit ( Ventana Medical Systems , Germany ) . Blood collected from STG90 and wild-type mice was centrifuged at 3 , 000 rpm for 10 min , and plasma was stored at −20°C until analysis . Plasma levels of total albumin , aspartate aminotransferase , alanine transaminase , and other total proteins were analyzed using an automatic blood chemistry analyzer ( Selectra II , Merck , Germany ) . All analyses were carried out using the GraphPAD Prism program ( GraphPAD Software , La Jolla , CA , USA ) . A one-way analysis of variance and Tukey's post hoc t-test were used for statistical analyses . Data are presented as mean ± SE . A p<0 . 05 was considered significant . The Kaplan-Meier survival analysis was used to compare survival against PRV infection . The statistical P value was generated between WT-PRV and STG90-PRV by the log-rank test . | Most strategies for developing virus-resistant transgenic cells and animals are based on the concept of virus-derived resistance , in which dysfunctional virus-derived products are expressed to interfere with the pathogenic process of the virus in transgenic cells or animals . However , these viral protein targeting approaches are limited because they only target specific viruses and are susceptible to viral mutations . We describe a novel strategy that targets the viral genome itself , rather than viral gene products , to generate virus-resistant transgenic cells and animals . We functionally expressed 3D8 scFv which has both DNase and RNase activities , in HeLa cells and transgenic mice . We found that the transgenic cells and mice acquired complete resistance to two DNA viruses ( HSV and PRV ) without accumulating the virus , and showed delayed onset of disease symptoms . The antiviral effects against DNA viruses demonstrated in this study were caused by ( 1 ) DNase activity of 3D8 scFv in the nucleus , which inhibited DNA replication or RNA transcription and ( 2 ) 3D8 scFv RNase activity in the cytoplasm , which blocked protein translation . This strategy may facilitate control of a broad spectrum of viruses , including viruses uncharacterized at the molecular level , regardless of their genome type or variations in gene products . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"biotechnology",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"molecular",
"cell",
"biology"
]
| 2014 | A Nucleic-Acid Hydrolyzing Single Chain Antibody Confers Resistance to DNA Virus Infection in HeLa Cells and C57BL/6 Mice |
Community-based public health campaigns , such as those used in mass deworming , vitamin A supplementation and child immunization programs , provide key healthcare interventions to targeted populations at scale . However , these programs often fall short of established coverage targets . The purpose of this systematic review was to evaluate the impact of strategies used to increase treatment coverage in community-based public health campaigns . We systematically searched CAB Direct , Embase , and PubMed archives for studies utilizing specific interventions to increase coverage of community-based distribution of drugs , vaccines , or other public health services . We identified 5 , 637 articles , from which 79 full texts were evaluated according to pre-defined inclusion and exclusion criteria . Twenty-eight articles met inclusion criteria and data were abstracted regarding strategy-specific changes in coverage from these sources . Strategies used to increase coverage included community-directed treatment ( n = 6 , pooled percent change in coverage: +26 . 2% ) , distributor incentives ( n = 2 , +25 . 3% ) , distribution along kinship networks ( n = 1 , +24 . 5% ) , intensified information , education , and communication activities ( n = 8 , +21 . 6% ) , fixed-point delivery ( n = 1 , +21 . 4% ) , door-to-door delivery ( n = 1 , +14 . 0% ) , integrated service distribution ( n = 9 , +12 . 7% ) , conversion from school- to community-based delivery ( n = 3 , +11 . 9% ) , and management by a non-governmental organization ( n = 1 , +5 . 8% ) . Strategies that target improving community member ownership of distribution appear to have a large impact on increasing treatment coverage . However , all strategies used to increase coverage successfully did so . These results may be useful to National Ministries , programs , and implementing partners in optimizing treatment coverage in community-based public health programs .
The health impact of community-based public health programs is dependent upon the proportion of the targeted population that is reached with the preventative health services , also known as treatment coverage . The World Health Organization ( WHO ) , country governments , and other institutions typically establish specific treatment coverage targets for community-based programs in order to benchmark programmatic success and achieve specific health outcomes . For example , the WHO recommends mass drug administration ( MDA ) treatment coverage with praziquantel and albendazole of at least 75% of children aged 6–15 years in schistosomiasis and soil transmitted helminth ( STH ) endemic areas . However , community-based public health programs often fail to achieve the pre-established coverage targets required to reduce morbidity , interrupt the transmission of diseases , or establish herd immunity [1] . Achieving high treatment coverage in the delivery of health services is critical to the success of community-based public health programs such as vitamin A supplementation ( VAS ) , child immunizations , and MDA campaigns targeting neglected tropical diseases ( NTDs ) . In order to reach large proportions of a population , these programs require strong public health platforms , typically decentralized outside of healthcare facilities . Delivery mechanisms include distribution through schools with or without the involvement of health professionals , delivery at community gathering points , or door-to-door distribution by health staff or volunteer community drug distributors ( CDDs ) . Challenges to achieving high treatment coverage include insufficient and inappropriate delivery systems , geographically remote populations , urban populations with high migration , competing resource and time priorities of distributors , and community misunderstanding or mistrust of the program [2–6] . For example , use of school-based delivery platforms to target school-aged children in areas with low school attendance rates compromises treatment coverage [7] . In Mali , Nigeria , and Sierra Leone , an insufficient number of volunteer drug distributors and low motivation among distributors was correlated with low MDA treatment coverage in lymphatic filariasis ( LF ) programs [2] . Additionally , community skepticism of LF MDA programs was associated with low participation of adults and adolescents in Tanzania . Mistrust was the result of community misunderstanding of the reason for distribution , a lack of information sharing , and fear of adverse side effects [3] . Given the multiple challenges to achieving high treatment coverage in community-based programs , it is important to identify and evaluate specific strategies that may have utility in increasing or sustaining high coverage . Sustaining high coverage is not only important for achieving health impact at the population level , but is also necessary within the context of a rights-based approach to health in order to ensure that individuals or groups are not repeatedly excluded from important health services . With this goal in mind , we sought to identify strategies that have been used to increase coverage in community-based programs and to evaluate their influence on treatment coverage . In doing so , we aim to generate evidence that may be useful to scaling community-based public health programs more broadly .
This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis ( PRISMA ) guidelines [8] . We systematically searched the electronic databases CAB Direct , Embase , and PubMed for articles that reported two or more treatment coverage estimates from community-based distribution of drugs , vaccines , or other public health services prior to and following an intervention intended to increase coverage . We used the search string ( “child health days*” OR chemoprevention OR “community based treatment” OR “community-directed” OR “community engagement” OR “deworm*” OR “mass drug administration” OR “preventative chemotherapy” OR “public health program” OR “supplementary immunization” OR “vaccin* campaign” OR “vitamin A” ) AND ( coverage OR compliance OR adherence ) AND “humans . ” No restrictions were made on study location , date of publication , article type , or language . Titles and abstracts were screened for all articles . All abstracts that described treatment coverage of a vaccine campaign , MDA , or other community-based distribution were included for full text review . If it was unclear if a study met inclusion criteria based upon the title and abstract , the text was read in full . According to pre-defined inclusion criteria , articles were excluded if they focused on facility-based distribution rather than community-based distribution . Articles that described supplemental or catch up campaigns were also excluded , as these represent a programmatic reaction to low coverage rather than a strategic intervention to improve coverage . The primary outcome of interest in this review was the measurable difference in treatment coverage of community-based public health distribution prior to and following implementation of interventions aimed at increasing coverage . Therefore , articles that qualitatively described barriers and/or methods to obtaining high coverage without reporting coverage estimates were excluded . Authors of included articles were contacted to clarify the study design or results , if needed . Article review and data abstraction were performed by a reviewer ( KVD ) and , in any circumstances where the eligibility of an article was unclear , two additional reviewers were consulted ( ARM and KHA ) . A structured data abstraction tool was used to abstract data from articles reviewed in full . Data abstracted from full texts included details on the intervention used to increase coverage , the size of the targeted and treated populations , dates of distribution , coverage estimates for baseline or control populations ( pre-intervention ) , and coverage estimates for follow up or experimental populations ( post intervention ) . Study design , data source , location , urban or rural setting of the distribution , and targeted disease were also collected . Individual treatment coverage estimates reported in articles were confirmed using reported number of targeted and treated individuals . If the 95% confidence interval ( CI ) for the coverage estimate was not reported , it was calculated using a one-sample proportion test and the sample size reported in the article . If the number of targeted or treated individuals was not reported in the article , it was calculated based upon the sample size and coverage estimates reported . If no sample size was reported , the corresponding author was contacted and requested to provide further details . Pooled coverage estimates for each strategy category were calculated by combining the sample sizes reported in each relevant study and using a one-sample proportion test to calculate proportion covered with 95% CIs . Percent change in treatment coverage for each study was calculated as the difference of pre- and post-intervention coverage estimates . All data analysis was conducted using the statistical computing software R ( R Foundation for Statistical Computing , Vienna , Austria ) and forest plots produced using the “forest plot” package [9 , 10] .
Six studies described MDA campaigns that were designed and delivered via a strategy of community-directed treatment ( ComDT ) compared to standard-of-care delivery [4 , 7 , 11–14] . The ComDT approach to delivery is characterized by efforts to systematically solicit community input in the organization and implementation of service delivery [36] . We have summarized these six studies briefly below . Nomadic communities in Nigeria are a hard-to-reach community for public health services . Akogun 2012 evaluated whether the ComDT approach could increase household ownership of insecticide-treated nets ( ITN ) compared to standard ITN distribution using a prospective study of experimental and control communities [11] . The experimental communities used ComDT to develop culturally acceptable methods for collecting and distributing ITNs , while the control communities continued to rely on standard distribution of ITNs . Coverage was defined as the percentage of households that owned an ITN . Each community had a baseline ITN coverage of zero . Comparing prospective cross sectional measures of ITN coverage after twelve months , Akogun 2012 found that ComDT distribution increased ITN coverage to 66 . 7% in the experimental community , while ITN coverage remained at zero in the control community ( Fig 2 ) . A community randomized controlled trial ( RCT ) in an urban setting in Orissa , India ( Babu 2006 ) evaluated the influence of ComDT delivery on treatment coverage in an LF MDA program [4] . Community participation in this study included establishing a steering committee of representatives from the local primary health center , local community-based organizations ( CBOs ) , NGOs , journalists , and government . The steering committee made key decisions concerning dates and locations of distribution , identifying community partners , selecting and training community drug distributors , and designing the accompanying social mobilization campaign . Standard-of-care MDA activities that did not systematically elicit community involvement were implemented in the control clusters . Coverage was calculated as the proportion of eligible community members who received treatment . ComDT delivery achieved 91 . 7% coverage compared to 68 . 6% coverage in control clusters ( Fig 2 ) . Gyapong 2000 used a community-RCT in four districts in Ghana to evaluate the influence of ComDT distribution on LF treatment coverage compared to standard MDA distribution [12] . In experimental communities , local health facility staff collaborated with community leaders to organize the MDA distribution method prior to drug distribution . Coverage was defined as the percent population who received treatment . Standard MDA was conducted in control communities . ComDT distribution achieved 73 . 8% treatment coverage , while standard MDA achieved 49 . 2% treatment coverage ( Fig 2 ) . The standard method for distribution of child health services in the Southern Province of Zambia is through government-run Child Health Weeks . Halwindi 2010 used a community-RCT to evaluate how ComDT could increase Child Health Week coverage [13] . Standard-of-care Child Health Week services were provided at health facilities and outreach posts in control sites . In experimental sites , services were distributed both at health facilities , as well as by community drug distributors through door-to-door and fixed-point distribution , as designed by the community . Coverage was defined as the percent of children aged 12 to 59 months who received services . Average treatment coverage achieved over four standard Child Health Weeks was 32 . 9% , while average treatment coverage achieved over four ComDT Child Health Weeks was 65 . 3% ( Fig 2 ) . A community-RCT in Tanzania compared the effect of a community-based ComDT approach to schistosomiasis and STH MDA delivery to standard-of-care school-based delivery ( Massa 2009 ) [7] . Villages selected their own community drug distributors , the drug delivery platform , and the timing of the campaign . Treatment coverage from two rounds of distribution was evaluated . The average change in treatment coverage from school-based to ComDT delivery among enrolled children was -1 . 5% , but 23 . 6% among unenrolled children ( Fig 2 ) . Wamae 2006 used a community-RCT in Kenya to compare the influence of ComDT on coverage of LF MDA as compared to standard MDA delivery [14] . Interactive trainings were conducted with health staff , community leaders , and community drug distributor volunteers on LF biology and the ComDT approach before drug distribution . Treatment coverage was defined as the percentage of people who received and swallowed tablets in each study arm . While standard LF MDA achieved 46 . 5% treatment coverage , ComDT achieved 88 . 0% treatment coverage ( Fig 2 ) . The six studies demonstrated a pooled increase in coverage of 26 . 2% , and an average absolute post-intervention coverage of 79 . 1% ( Fig 2 ) . Building upon the concept of ComDT , Katabarwa 2010 evaluated if MDA treatment coverage is higher in areas where people tend to live near relatives due to the leveraging of kinship networks [15] . Kinship-networks utilize the same approach as ComDT distribution , however the distribution is conducted within kinship groups . This approach requires more drug distributors than the standard approach as distribution is targeted to smaller groups . Coverage was defined as the percent population who received treatment for onchocerciasis . While ComDT alone achieved an average treatment coverage of 69 . 3% over two years , kinship-enhanced distribution achieved an average treatment coverage of 93 . 7% over that same period ( Fig 2 ) . Calderon-Ortiz 1996 used a community-RCT in Mexico to evaluate the influence of permanently hiring community vaccine distributors on the percentage of children less than one year of age who complete their vaccine schedules [17] . This was contrasted with standard vaccine campaigns distributed by temporary vaccine distributors . Coverage was defined as the percent of children less than one year in age who completed their vaccine schedule . While the standard vaccine campaign resulted in 21 . 1% coverage , permanently hiring community distributors to provide vaccines resulted in 93 . 5% coverage ( Fig 2 ) . Muhmuza 2013 used prospective repeat cross-sectional measures to evaluate the influence of drug distributor incentives on schistosomiasis MDA treatment coverage in 12 primary schools in Uganda [16] . Incentives for the teachers and health staff that distributed drugs consisted of a one-day refresher course on drug distribution and t-shirts printed with “NTD Control Program . ” Coverage was calculated as the proportion of targeted children who received treatment . Standard-of-care distribution without distributor incentives resulted in coverage of 28 . 2% of children while the incentives strategy implemented the following year resulted in coverage of 48 . 9% ( Fig 2 ) . The pooled increase in treatment coverage from these two studies was 25 . 3% , with an average absolute post-intervention coverage of 52 . 9% ( Fig 2 ) . Eight studies used intensified IEC activities , including the development and distribution of posters , flyers , leaflets , brochures , and radio and TV broadcasting of health messaging in the local language to promote awareness of distribution activities , provide health education , and promote behavior change , to increase treatment coverage [5 , 18–24 , 37] . Coverage of each MDA program was calculated as the proportion of the total targeted population that received treatment . A community-RCT in Orissa State , India evaluated the influence of intensified IEC activities on LF MDA treatment coverage when implemented in the month preceding MDA ( Cantey 2010 ) [18] . Compared to the standard pre-MDA education campaign promoted by MOH , IEC activities in experimental clusters consisted of murals , street plays , radio and newspaper advertisements , village education sessions , and increased dissemination of posters and leaflets . The IEC materials described LF transmission mechanisms , drug side effects , and mosquito control measures . Experimental clusters achieved 59 . 5% MDA treatment coverage compared to 52 . 2% in control clusters ( Fig 2 ) . Habib 2017 used a community-RCT to evaluate the influence of an IEC campaign and integrated distribution on OPV coverage in Pakistan [19] . The control arm used routine vertical immunization services , while the experimental arm used integrated distribution of OPV and maternal and child health services preceded by IEC activities . IEC materials included information on maternal health services , nutrition , hygiene and sanitation , and OPV . Community mobilizers distributed IEC materials with the aid of pictorial booklets and counseling cards for reference . Coverage was defined as the percent of children aged 1 month to five years who received OPV . In the control arm , where OPV was distributed vertically with no accompanying IEC , OPV treatment coverage was 75% . In the experimental arm , where OPV was integrated with other services and incorporated IEC materials , OPV treatment coverage was 82% ( Fig 2 ) . A study in American Samoa used prospective repeat cross-sectional sampling to evaluate IEC-associated changes in LF MDA coverage ( King 2011 ) [20] . The intensified behavior change campaign consisted of brief radio and television broadcasts of skits , testimonials , and announcements describing LF transmission and prevention , and locations of MDA distribution . Broadcasts were made during morning shows , before local news broadcasts , and during high profile sporting events prior to and during MDA . Treatment coverage increased from 49 . 6% at baseline to 71 . 0% at follow up ( Fig 2 ) . The study also transitioned MDA from door-to-door delivery method to fixed-point delivery , as described in the following section . A community-RCT in urban Kenya evaluated the effect of an enhanced IEC campaign on LF MDA treatment coverage in an area with historically low campaign participation ( Njomo 2014 ) [5] . In experimental clusters , MDA was preceded by increased dissemination of posters and banners in public places , announcements made by loudspeaker on the day before and day of distribution , and increased announcements made from mosques , churches , and schools . Experimental clusters that implemented the enhanced IEC campaign achieved 72 . 2% coverage , compared to 70 . 4% in control clusters conducting standard IEC activities ( Fig 2 ) . In Akre District , Iraq , an area with historically low childhood vaccine coverage , Rahman 2012 used a prospective cross-sectional study to evaluate how implementation of an IEC campaign that included messaging from community religious leaders influenced vaccine coverage [21] . Coverage was defined as the percent of children less than one year in age who received vaccines . Standard vaccine distribution at baseline was conducted with no accompanying education program one to three times per week at health clinics or by mobile outreach teams , achieving 36 . 9% coverage of children less than one year of age . When the research team implemented an IEC program preceding vaccine distribution , including announcements over loudspeakers , health talks , posters , and video , vaccine coverage increased to 87 . 6% ( Fig 2 ) . A prospective repeat cross-sectional study in India compared LF MDA treatment coverage after an enhanced communication campaign to standard-of-care ( Ramaiah 2006 ) [22] . The communication campaign consisted of increased distribution of registration slips , posters , and badges compared to the standard-of-care . The campaign also introduced novel IEC activities such as distribution of ribbon flags , leaflets , t-shirts , cotton bags , and malted drink sachets , in addition to intensive television advertising . The IEC campaign also involved CDDs , government at the state , district and village levels , and a touring bicycle team wearing “Filaria Day” t-shirts to advertise the distribution day . The enhanced campaign resulted in 88 . 0% treatment coverage compared to 71 . 0% coverage observed at study baseline ( Fig 2 ) . While the primary aim of de Rochars’s ( 2005 ) study in Haiti was to demonstrate the effect of annual LF MDA on microfilaremia prevalence over three rounds of treatment , the study also evaluated the influence of a communication campaign on treatment coverage [23] . Despite extensive community sensitization at baseline , treatment coverage decreased during the second round of MDA . Therefore , prior to the third treatment round , researchers implemented a communication campaign consisting of radio broadcasts by community leaders and televised videos about benefits and risks of LF preventive chemotherapy . The communication campaign was associated with an increase in coverage from 53 . 0% to 80 . 8% ( Fig 2 ) . Using a prospective cross-sectional study , Zimicki 1994 evaluated the influence of a national mass media communication campaign on full child vaccine coverage in the Philippines [24] . Prior to 1990 , standard childhood vaccine distribution occurred in health centers with no accompanying IEC campaigns . In 1989 , 50 . 7% of children were fully covered by standard vaccine campaigns . The mass media IEC campaign implemented in 1990 popularized a single day of the week as “vaccine day , ” and provided information on measles vaccines in particular , as measles was a threat well known by mothers . The communication campaign resulted in 60 . 4% of children completing their vaccine schemes on time ( Fig 2 ) . The pooled increase in treatment coverage from these eight studies was 21 . 6% , with an average absolute post-intervention coverage of 77 . 8% ( Fig 2 ) . In addition to intensifying IEC activities , King 2011 ( described above ) evaluated the influence of changing the distribution method for LF MDA [14] . Standard-of-care distribution was conducted door-to-door by health staff , achieving 49 . 6% treatment coverage . The following year , MDA was delivered in fixed-point community spaces including churches , schools , places of employment , shopping centers , bingo halls , and the airport . This change was informed by distributor and community feedback on the feasibility and cultural appropriateness of distribution platform . Post-intervention coverage was 71 . 0% ( Fig 2 ) . Conversely , Linkins 1995 used a prospective cross-sectional study to evaluate OPV coverage in Egypt using door-to-door distribution as compared to standard-of-care fixed-point delivery . Coverage was defined as the percent of children less than five years of age who received OPV . Standard-of-care fixed-point distribution was associated with a coverage of 86 . 0% while door-to-door distribution was associated with 100% coverage [25] ( Fig 2 ) . Although overall program costs were higher with the more intensive doo-to-door distribution method , cost per child was comparable to fixed-point delivery given the higher coverage achieved . Nine studies included in this review evaluated the impact on coverage of integrating distribution of community-based public health services [19 , 26–33] . One such platform for integration is Child Health Days , or Child Health Weeks or Months , which are implemented routinely throughout Eastern and Southern Africa as a strategy for efficiently and comprehensively distributing packages of key preventative health services to children such as immunizations , nutritional supplementation , and insecticide-treated bed nets [38] . In Central Nigeria , Blackburn 2006 used a prospective cross-sectional study to evaluate how integration of ITN distribution with an existing LF and onchocerciasis MDA program would affect the coverage of ITN ownership [26] . Coverage was defined as the percent of children less than 5 years of age and pregnant women who owned an ITN . ITNs were distributed simultaneously with LF and onchocerciasis MDA . Prior to integrated service distribution , ITN coverage was 9 . 0% . ITN coverage increased to 74 . 0% when distributed in conjunction with the existing MDA program ( Fig 2 ) . In 2009 , a nationally integrated MDA program targeting LF , onchocerciasis , trachoma , schistosomiasis , and STH was implemented in Mali [27] . Dembele 2012 used a prospective cross-sectional study to compare program coverage achieved by national disease-specific vertical distribution in 2006 to that achieved by national integrated distribution in 2009 . Program coverage was defined as the proportion of the population treated among the eligible population in targeted areas . Integrated distribution increased treatment coverage for schistosomiasis and STH from 40 . 1% to 71 . 4% , and from 40 . 5% to 91 . 5% , respectively . Onchocerciasis treatment coverage increased only slightly , from 101 . 6% to 103 . 8% . However , LF and trachoma treatment coverage decreased from 97 . 6% to 91 . 5% and 104 . 9% to 78 . 5% , respectively . Overall , treatment coverage increased 19 . 3% with integrated service distribution ( Fig 2 ) . Two studies ( Doherty 2010 and Oliphant 2010 ) used retrospective data reported in Demographic and Health Surveys ( DHS ) from Ethiopia , Madagascar , Tanzania , Uganda , Zambia , and Zimbabwe to evaluate how coverage of vitamin A supplementation and measles immunization campaigns were influenced by integrating these services in national Child Health Days [28 , 33] . Data abstracted from DHS included the number of children age 6–59 months who received vitamin A supplementation in the 6 months preceding the survey and the number of children age 12–23 months who received a measles vaccine at any time prior to the survey . Treatment coverage was calculated as the proportion of children sampled that received services . The change in coverage was calculated by subtracting the most recent coverage estimate from the DHS report preceding integration from the DHS estimate following integration . Prior to implementation of CHDs , vitamin A supplementation coverage in Ethiopia , Madagascar , and Tanzania was 55 . 8% , 4 . 2% and 13 . 8% , respectively . Following integration into the CHD platform , coverage in Ethiopia decreased to 45 . 8% , but in Madagascar and Tanzania increased to 76 . 2% and 45 . 5% , respectively . The effect of integrating measles immunization programs into the CHD structure was also mixed . Coverage in Tanzania and Zambia did not change significantly following CHD implementation , while coverage in Zimbabwe decreased from 79 . 1% to 65 . 6% . However , coverage in Ethiopia and Uganda increased from 26 . 6% and 56 . 8% to 34 . 9% and 68 . 1% , respectively . While the effect of integrated distribution on coverage in each country varied , Doherty 2010 found that average coverage increased 47 . 1% , with an average absolute post-intervention coverage of 55 . 5% , and Oliphant 2010 found that average coverage decreased 0 . 8% with an absolute post-intervention coverage of 50 . 0% ( Fig 2 ) . In 2007 , a nationwide integrated distribution campaign of measles vaccines , deworming drugs , and vitamin A supplementation targeting children less than 59 months of age was launched in Madagascar [29] . ITN distribution was included in the campaign in 59 districts . Goodson 2012 compared cross-sectional measures of measles vaccine coverage in areas that were and were not targeted for ITN distribution . Coverage was defined as the percentage of children less than 59 months who received a measles vaccine through the campaign . In areas without concurrent ITN distribution , measles vaccine coverage was 61 . 5% while in areas with concurrent ITN distribution coverage was 71 . 0% ( Fig 2 ) . Conversely , Grabowsky 2005 used a prospective cross-sectional study in five districts in Zambia to evaluate the influence of integration of ITN delivery within national measles vaccine campaigns [30] . Coverage was defined as the percentage of children aged 9 months to 14 years who received an ITN during a one-week measles vaccine campaign . Preceding integrated distribution , ITN coverage was 22 . 7% . Integrated delivery with the vaccine campaign achieved ITN coverage of 80 . 2% ( Fig 2 ) . The previously described study by Habib 2017 evaluated the influence of integrated distribution in Pakistan on OPV treatment coverage [19] . The control arm used routine vertical immunization services , while the experimental arm used integrated distribution of OPV and maternal and child health services preceded by IEC activities . In the control arm , where OPV was distributed vertically with no accompanying IEC , OPV treatment coverage was 75% . In the experimental arm , which integrated distribution of OPV with other services and incorporated IEC materials , OPV treatment coverage was 82% ( Fig 2 ) . Mwingira 2016 used a prospective cross sectional study to evaluate the influence of integrating a measles and rubella vaccine campaign with MDA for onchocerciasis and LF in Tanzania on MDA treatment coverage compared to vertical distribution [31] . In addition to integrating distribution of the vaccine and MDA , planning exercises , community sensitization , media campaigns , and monitoring and evaluation for both the vaccine campaign and MDA were integrated . Coverage was defined as the percent population who received MDA . Vertical distribution of the MDA resulted in 86% treatment coverage , while integrated distribution resulted in 93% coverage ( Fig 2 ) . As previously described , Ndyomugyenyi 2003 evaluated how integration of schistosomiasis and STH treatment with existing community-directed onchocerciasis treatment influences onchocerciasis treatment coverage [32] . STH treatment was distributed to children aged 5 to 14 years , and onchocerciasis treatment was distributed to all eligible community members . Coverage was defined as the percentage of the eligible population who received onchocerciasis treatment . Integration improved onchocerciasis treatment coverage from 77 . 2% to 81 . 3% over standard community-directed treatment ( Fig 2 ) . The pooled change in coverage from these nine studies demonstrated a 12 . 7% increase in coverage , and an average absolute post-intervention coverage of 90 . 0% ( Fig 2 ) . Three articles evaluated changes in MDA treatment coverage when delivered as community-based campaigns as opposed to standard school-based programs [7 , 32 , 34] . School-based programs are characterized by distribution of health services to school-aged children by teachers and/or health staff in schools . Children who are not enrolled in schools are typically invited to attend treatment days to receive the health service , and adults are not targeted for treatment . In these studies , community-based programs combined school-based distribution to target children enrolled in school as well as distribution in community spaces to target unenrolled children and adults . In both studies treatment coverage was calculated as the proportion of targeted individuals who received treatment . The previously described study in Tanzania ( Massa 2009 ) compared coverage of standard-of-care school-based MDA to community-based MDA for schistosomiasis in enrolled and unenrolled school-age children [7] . Two rounds of community-based delivery were implemented and coverage was estimated . Among enrolled school-aged children , the first round of the intervention achieved similar coverage to the standard-of-care ( 80 . 3% compared to 82 . 1% , respectively ) . Among unenrolled children , however , the intervention achieved higher coverage than the standard-of-care ( 80 . 0% compared to 59 . 2% , respectively ) . Similar results were observed following a second round of the intervention; coverage of children enrolled in school was not significantly different from the standard-of-care among enrolled children , whereas coverage amongst unenrolled children increased relative to the standard-of-care among unenrolled children ( 80 . 7% in intervention cluster compared to 78 . 4% in standard-of-care cluster ) ( Fig 2 ) . In Arua District , Uganda , Ndyomugyenyi 2003 used a community-RCT to compare distribution of schistosomiasis and STH treatment through community-based distribution and school-based distribution [32] . Coverage was defined as the percentage of children aged 5 to 14 years who received treatment . School-based distribution by teachers in primary school was associated with coverage of 79 . 0% , while community-based delivery by community drug distributors was associated with treatment coverage of 85 . 0% ( Fig 2 ) . Another study ( Oshish 2011 ) used prospective repeat cross-sectional measures to evaluate changes in schistosomiasis MDA treatment coverage of school-age children ( 6–18 years ) and adults when transitioning from standard-of-care school-based MDA to community-based MDA [34] . Standard-of-care delivery had 94 . 0% coverage of enrolled school-aged children , 68 . 0% coverage of unenrolled school-aged children , and did not target adults . Community-based delivery via schools , health centers , markets , mosques , and community leaders’ residences achieved 97 . 9% coverage of enrolled children , 90 . 0% coverage of unenrolled children , and 73 . 9% coverage of adults , with an overall coverage of 82 . 5% ( Fig 2 ) . The pooled change in coverage from these three studies demonstrated an 11 . 8% increase in coverage , and an average absolute post-intervention coverage of 94 . 5% among school-aged children ( Fig 2 ) . One study ( Ladner 2014 ) evaluated the influence of campaign management source on the treatment coverage in community-based public health programs [35] . The study used linear regression to retrospectively evaluate factors that were associated with Gardasil vaccine coverage among girls age 9 to 13 years in several low- and middle-income countries using data from the Gardasil Access Program ( GAP ) . Coverage was calculated as the proportion of targeted girls who received the full vaccine course . Distribution management by an NGO compared to an MOH was found to be significantly associated with higher vaccine coverage ( p = 0 . 05 ) . Across 14 countries , programs under NGO management achieved 93 . 1% coverage , while programs under MOH management achieved 87 . 3% coverage , not accounting for distribution duration or community involvement ( Fig 2 ) .
Delivering community-based public health interventions at scale can be challenging and impact is often limited by the ability to reach targeted populations . Identifying strategies that can improve treatment coverage is relevant for a number of ongoing and planned interventions . In this review , we identify several strategies that appear to demonstrate a positive impact on the treatment coverage in community-based public health programs . Strategies that increased the community’s role in and ownership of distribution events had the largest positive influence on change in treatment coverage . In contrast , strategies that achieved the highest absolute coverage did so by increasing access to services through door-to-door and community-based distribution . In areas with insufficient coverage , we suggest that public health officials and implementation partners consider the ‘supply’ and ‘demand’ side factors influencing community member participation in programs . For example , if programs are not reaching targeted populations with services than they may consider transitions in delivery strategy ( ex . door-to-door delivery vs . fixed point delivery ) in order to increase coverage . On the other hand , if coverage is low because community members choose not to comply , then programs may wish to consider strategies such as community-directed treatment or intensified IEC campaigns . This study provides a menu of potential strategies from which public health officials and partners can evaluate which coverage interventions may be most appropriate within their own setting . Additional analyses to determine the cost-effectiveness and feasibility of each strategy , or combinations of strategies , are needed to inform program and policy decisions to optimize these programs at scale . These findings are particularly important for studies that aim to interrupt the transmission of disease by providing preventative interventions with extremely high coverage . | Many public health platforms provide decentralized interventions outside of health facilities , including mass drug administration for neglected tropical diseases , immunizations , vitamin supplementation , and others . The purpose of these community-based public health platforms is to reach large proportions of populations in need with important preventative healthcare . However the platforms require high treatment coverage of targeted populations in order to achieve health impact . And , in many cases , targeted populations are low-income , rural , and hard to reach with large health campaigns . The purpose of this systematic review is to evaluate strategies for achieving high treatment coverage in public health service distribution programs . We identified nine different strategies used to increase coverage of distribution programs . Community-directed distribution was associated with the largest increase in treatment coverage . Similarly , incentivizing distributors also had a strong influence on increasing treatment coverage . These findings have important implications for governments , implementers , and funders who aim to provide health services at scale . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
]
| [
"children",
"medicine",
"and",
"health",
"sciences",
"tropical",
"diseases",
"parasitic",
"diseases",
"pediatrics",
"vaccines",
"age",
"groups",
"neglected",
"tropical",
"diseases",
"infectious",
"disease",
"control",
"onchocerciasis",
"families",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"agrochemicals",
"child",
"health",
"agriculture",
"insecticides",
"people",
"and",
"places",
"helminth",
"infections",
"schistosomiasis",
"measles",
"biology",
"and",
"life",
"sciences",
"population",
"groupings",
"viral",
"diseases"
]
| 2018 | Strategies to improve treatment coverage in community-based public health programs: A systematic review of the literature |
Dengue-suppressing Wolbachia strains are promising tools for arbovirus control , particularly as they have the potential to self-spread following local introductions . To test this , we followed the frequency of the transinfected Wolbachia strain wMel through Ae . aegypti in Cairns , Australia , following releases at 3 nonisolated locations within the city in early 2013 . Spatial spread was analysed graphically using interpolation and by fitting a statistical model describing the position and width of the wave . For the larger 2 of the 3 releases ( covering 0 . 97 km2 and 0 . 52 km2 ) , we observed slow but steady spatial spread , at about 100–200 m per year , roughly consistent with theoretical predictions . In contrast , the smallest release ( 0 . 11 km2 ) produced erratic temporal and spatial dynamics , with little evidence of spread after 2 years . This is consistent with the prediction concerning fitness-decreasing Wolbachia transinfections that a minimum release area is needed to achieve stable local establishment and spread in continuous habitats . Our graphical and likelihood analyses produced broadly consistent estimates of wave speed and wave width . Spread at all sites was spatially heterogeneous , suggesting that environmental heterogeneity will affect large-scale Wolbachia transformations of urban mosquito populations . The persistence and spread of Wolbachia in release areas meeting minimum area requirements indicates the promise of successful large-scale population transformation .
Dengue fever is the most common arboviral disease affecting humans [1] . Over 2 , 500 , 000 , 000 people live in dengue-afflicted regions , and dengue incidence is increasing at an alarming rate in tropical and subtropical countries [2] . A number of other arboviruses also represent emerging disease risks , including chikungunya and Zika , the latter being associated with a recent explosive epidemic in South America [3 , 4] . The main approach to controlling these diseases has been suppression of the principal mosquito vector , Ae . aegypti , either through source reduction or insecticide-based control programs . Given the increasing incidence of Ae . aegypti-associated human disease , it is clear that current control measures are insufficient . In response to this problem , a number of new control approaches are currently being developed and tested [5 , 6 , 7 , 8 , 9 , 10] . In contrast to control efforts that require repeated population suppression , the Eliminate Dengue Program ( http://www . eliminatedengue . com/program ) aims to modify populations using long-lasting local introductions of a dengue-inhibiting Wolbachia into naturally uninfected populations of Ae . aegypti . The strain , wMel , was transferred from Drosophila melanogaster into laboratory-raised Ae . aegypti , who inherit the infection maternally [11 , 12] . Following introgression of the infection into a native genetic background , Wolbachia-infected mosquitoes are released into the field to mate with wild uninfected mosquitoes , and wMel frequency increases through cytoplasmic incompatibility ( CI ) [13 , 14] . CI describes the fact that uninfected females mated with Wolbachia-infected males produce inviable embryos . In Ae . aegypti , this is believed to occur in 100% of these incompatible crosses [11] . In contrast , infected females can mate with either infected or uninfected males and produce almost 100% infected progeny . CI greatly reduces the relative fitness of uninfected females when infected males are common and drives rapid establishment of Wolbachia in isolated mosquito populations [14] , given that there is no mating bias against wMel-infected Ae . aegypti [15] . Although wMel-infected females receive a frequency-dependent relative fitness advantage from CI , they also suffer from frequency-independent fitness costs , including decreases in fecundity and larval competitive ability [16 , 17 , 18 , 19] . Thus , CI does not produce a net fitness advantage while wMel is rare , resulting in dynamics analogous to those produced by an Allee effect in ecology [20 , 21] and by natural selection on a locus ( or alternative karyotypes ) in which heterozygotes are less fit than either homozygotes ( i . e . , underdominance , [22 , 23 , 24] ) . The interaction of the frequency-dependent advantage associated with CI and the frequency-independent cost ( s ) produces “bistable dynamics” with a threshold frequency of infection ( denoted p^ ) below which the infection will be locally eliminated and above which frequencies systematically increase [25 , 26 , 27] . Curtis [23] first proposed transforming pest populations by introducing translocations that are expected to show bistable dynamics ( cf . [28] ) . The bistable model for Wolbachia spread was introduced by Turelli and Hoffmann [29] to explain the rapid spread of wRi , a CI-causing Wolbachia variant , through California populations of Drosophila simulans . Although this interpretation of wRi dynamics has now been challenged by more recent data on the spread of natural Wolbachia infections [30] , 3 lines of evidence nevertheless support bistability of wMel-transinfected Ae . aegypti [31]: ( 1 ) frequency dynamics from the original field releases [14] , ( 2 ) direct experimental evidence for lower fecundity and viability [19 , 32] , and ( 3 ) new data showing that persistent influx over 2 years of wMel-infected Ae . aegypti into a relatively isolated population has not led to establishment of wMel there [31] . In order for the invasion to spread spatially under bistability , new uncolonised areas must receive infected immigrants at a rate high enough to be pushed past the threshold frequency , p^ . Under the dynamics produced by CI-inducing Wolbachia , spatial spread is expected in a habitat with relatively homogeneous population densities if p^ is below a critical value near 0 . 5 [20 , 29] . For wMel in Ae . aegypti near Cairns , p^ is thought to be moderate ( p^≈0 . 2–0 . 35 ) because of its relatively low fitness costs and near-perfect maternal transmission [11 , 14 , 31] . Previously , wMel-infected Ae . aegypti released in 2 relatively isolated communities in Northern Queensland , Australia ( Gordonvale and Yorkeys Knob ) , colonised each area rapidly [14] , and the infection has persisted at high frequency ( >90% ) at both sites [18] . Moreover , wMel continues to show strong blockage of dengue transmission in laboratory-challenged mosquitoes derived from field collections [33] . Here we present data from 3 subsequent releases of wMel-infected Ae . aegypti in Cairns , Northern Queensland , a city with about 150 , 000 residents that is located between the communities of Gordonvale and Yorkeys Knob . These releases followed protocols similar to those of [14] , but the release zones were centred within suburban landscapes , providing a continuous habitat for Ae . aegypti . This study investigates the capability of the wMel infection to spread spatially through urban Ae . aegypti populations and the stability of the infection in invaded regions over time . Spread from localized releases to surrounding uninfected areas depends on mosquito dispersal and relative population densities . Spatial spread can be slowed or stopped if densities are higher in surrounding uninfected areas [20] . Dispersal of Ae . aegypti varies with local environmental conditions . Poor habitats generally induce larger dispersal distances as gravid females must travel further to find the relatively rare oviposition sites [34 , 35 , 36] . Despite its global success as an invasive species in tropical habitats , presumably through dispersal of eggs and larvae [37] , adult Ae . aegypti are generally considered weak dispersers . Females usually remain within 50–150 m of their eclosion site [34 , 38 , 39 , 40 , 41 , 42] . They appear to disperse poorly across highways [31 , 42 , 43] and through vegetated parkland [44] . Occasional long-range dispersal , on the order of 0 . 5–1 km , has been observed [45 , 46 , 47 , 48] . However , given the bistable dynamics of wMel in Ae . aegypti , rare long-range dispersal will not accelerate Wolbachia spread because the infection will not increase locally from low initial frequencies [20 , 31] . We document local wMel establishment and heterogeneous spatial spread from the 2 relatively large release areas . Our new data demonstrate that local Wolbachia introductions can succeed , persist for at least 2 years , and produce slow spatial spread . Using graphical summaries , we approximate the rate of spatial spread and the width of the spreading wave . We also show that our field data are broadly consistent with simple mathematical models that depend critically on bistable frequency dynamics for wMel transinfected into Ae . aegypti . These models involve only 2 parameters , one describing the position of the unstable threshold point , p^ , and the other , σ , describing average Ae . aegypti dispersal distance . Both parameters can be estimated independently of spread data [31] . We also present likelihood-based data analyses that fit simple curves to estimate the shape and speed of Wolbachia spread . The shape of the advancing wave is summarized by wave width , defined as the inverse of the maximum slope in infection frequencies , averaged over the wave front [49] . As discussed below , wave width provides an estimate of dispersal distance averaged over time . Wave speed is defined as the average rate of movement of an intermediate infection frequency ( e . g . , 0 . 5 . ) The theory of bistable waves leads to a simple prediction for wave speed in terms of wave width and p^ , the threshold infection frequency above which local increases in infection frequencies ( p ) are expected [20 , 24 , 31 , 50] . The observed speed of wMel spread in Cairns is broadly compatible with this prediction , and the estimated wave width is also consistent with independent estimates of dispersal . Moreover , the lack of clear establishment or spread from our third , significantly smaller , release area ( only 0 . 11 km2 ) is consistent with the prediction for bistable dynamics that releases must be conducted over sufficiently large areas to initiate spatial spread . Our likelihood analyses also quantify significant heterogeneity in rates of spatial spread that is apparent from our graphical representations . We attempt to link this heterogeneity to easily measured habitat variables . Heterogeneity in host population density is expected to strongly influence Wolbachia invasions subject to bistable dynamics , especially affecting wave speed and potentially restricting the extent of spread [20] . Even if Ae . aegypti disperse equally in all directions , heterogeneities in population density produce asymmetries in net migration . This asymmetry accelerates spread from high-density patches to low-density patches and decelerates—or halts—spread out of low-density patches [20] . Habitat variables such as shade , yard condition , and abundance of oviposition sites have been correlated with Ae . aegypti abundance [41 , 51]; the frequency of Wolbachia infection within the release zone in Gordonvale , Queensland , was higher in neighbourhoods with more brick and screened houses , which are associated with lower Ae . aegypti abundance [32] . This motivates our attempts to understand patterns of local spread by inferring local densities from easily measured habitat variables . However , the variables we assessed did not predict observed heterogeneities in spread beyond the release zones .
Fig 1 shows the 3 areas in Cairns , Queensland , where Eliminate Dengue staff released Ae . aegypti adults infected with the wMel strain of Wolbachia between January 10th and April 24th , 2013 . The release zones , located in the suburbs of Edge Hill/Whitfield ( EHW ) , Parramatta Park ( PP ) , and Westcourt ( WC ) , were within 2 km of each other and encompassed 0 . 97 km2 , 0 . 52 km2 , and 0 . 11 km2 , respectively . Mosquitoes were released evenly throughout each release zone at weekly intervals . Total BG-Sentinel trap collections for EHW , PP , and WC are summarised in S1 Table . Our collections continued for about 2 years and are summarized in 4 time intervals . The first dry season D1 ( May 2013–October 2013 ) , began immediately after the releases , followed by the first wet season W1 ( November 2013–April 2014 ) , the second dry season D2 ( May 2014–October 2014 ) , and the second wet season W2 ( November 2014–April 2015 ) . Weekly trap yields at EHW and PP decreased progressively from the onset of each dry season but rose again sharply at the beginning of each wet season ( Fig 2 ) . Mosquitoes were caught in consistently higher numbers at PP than at EHW ( two-tailed Student t test: P < 0 . 001 ) , and onsite traps ( traps within the release zone ) collected mosquitoes at a faster rate than offsite traps ( traps outside the release zone; two-tailed Student t test: P < 0 . 001 ) . When accounting for seasonal changes , yields of uninfected mosquitoes caught in offsite traps at both sites tended to decrease over time ( Fig 2 panel A ) . At PP , there was a corresponding increase in infected mosquito numbers , while at EHW , infected mosquito yields were relatively consistent throughout . Among onsite traps , yields of infected mosquitoes were consistent with seasonal expectations and were stable over time ( Fig 2 panel B ) . The higher local infection frequencies onsite might lead to an assumption that uninfected mosquito numbers would decline more rapidly than those offsite , but this was not observed , with uninfected mosquito yields at both sites increasing sharply in W2 . This proliferation was particularly surprising considering the 2-month-long periods at EHW in the previous season , D2 , during which no uninfected mosquitoes were caught onsite ( Fig 2 panel B ) . EHW and PP were both invaded quickly , and by the time releases had finished , Wolbachia infection frequencies within each release zone had reached p = 0 . 85 ( S1 Fig ) . Following the final releases , p remained relatively stable and near fixation within each release zone . However , in W2 , onsite p at EHW dropped from 0 . 96 to 0 . 84 , the lowest recorded since monitoring began . Considering Fig 2 panel B , it appears that this was due to neither imperfect maternal transmission of Wolbachia [26] nor increased mortality among infected mosquitoes , as their numbers increased to levels similar to those observed in W1 . Rather , a sudden influx of uninfected mosquitoes seems most plausible . Averaged across all 4 seasons , 0 . 88 of mosquitoes within the EHW release zone were infected , while 0 . 90 were infected at PP . The WC release zone was invaded as quickly as EHW and PP . However , beginning in September 2013 , onsite p dropped sharply to p < 0 . 7 , after which frequencies fluctuated . While onsite p never dropped below any plausible value for p^ , at no point did the invasion at WC exhibit either the near-fixation values of p or the temporal stability observed at both EHW and PP ( see figures below and compare panel C of S2 Fig with panels A and B ) . The changes of p with time at EHW , PP , and WC between 7 May 2013 and 30 April 2015 are displayed in Figs 3 , 4 and 5 , respectively , along with trap locations and yields . The plots , based on spatial averaging ( ordinary Kriging as described in the Methods section , performed using ArcMap 10 . 2 . 2 [52] ) , show considerable seasonal heterogeneity in the spatial structure of the invasions at EHW , PP , and WC . At EHW ( Fig 3 ) after D1 , the infection was confined largely to the north and northeast , but by the end of W1 , the invasion had spread to the east , northeast , and southwest . This pattern persisted through D2 , with a small retraction in the north and expansion in the east , though for this season , Kriging was affected by a small sample size ( N = 31 ) . Kriging on W2 trap data demonstrated 3 main shifts from this pattern: the continued expansion to the north , northeast , and east; the successful invasion of the west; and the apparent reduction in p from p ≥ 0 . 8 to 0 . 65 ≤ p ≤ 0 . 8 at 3 traps in the centre of the release zone . At PP ( Fig 4 ) , spread through D1 was confined mostly to the southeast , from the edge of the release zone up to Mulgrave Road . In the following season ( W1 ) , infected mosquitoes were found south across Mulgrave Road and north of the release zone . The infection persisted south of Mulgrave Road but only at below-threshold ( p ≤ 0 . 3 ) frequencies . Over D2 , the invasion expanded in range , with high frequencies observed in the north and the southeast and moderate frequencies in the northwest . At both EHW and PP , the area covered by the infection tended to increase over time ( Figs 3 and 4; summarized in panels A and B of S2 Fig ) , except for PP in W1 , in which the area within the p ≥ 0 . 8 contour decreased by 3% from D1 , and for EHW in D2 , in which the area within the p ≥ 0 . 8 and p ≥ 0 . 5 contours decreased by 7% and 1% , respectively . Nevertheless , from D1 to W2 , the area enclosed by the p ≥ 0 . 8 contours grew by 85% at EHW and 77% at PP . At WC ( Fig 5; S2 Fig panel C ) , establishment or spread of the infection was not observed . Following D1 , the Wolbachia invasion failed to expand to the south or west of the release zone . Within the release zone , a gradual retreat of the p ≥ 0 . 8 contour was observed , with several onsite traps in W2 registering p ≤ 0 . 3 . The area covered by the infection at WC reached a peak at W1 , but by W2 the area enclosed by the p ≥ 0 . 8 and p ≥ 0 . 5 contours had decreased by 52% and 44% , respectively , from this maximum ( S2 Fig panel C ) . Wolbachia also failed to spread from WC into PP ( or vice versa ) , but these areas are separated by parkland , which is likely to act as a barrier to movement and prevents ongoing monitoring there . The time interval from D1 to W2 is around 1 . 5 years , which can be approximated as 15 generations , assuming about 10 generations per year ( explained below ) , or simply viewed as 548 days . When containers are initially colonised and food is available for larvae , developmental time is likely to be rapid at 7–10 days . However , larval populations can rapidly exceed the carrying capacity of the container and its food source ( typically leaves ) , and development is then slowed to 20–50 days [53]; these variable conditions produce a range of adult body sizes that is typically found in field samples from Cairns [54] . If we assume an intermediate value of 20 days in the field , along with time for adult maturation to mating and blood feeding ( 2–3 days post eclosion ) , blood-meal digestion and egg formation and oviposition ( 4 days ) , and egg embryonation ( 3 days ) [55] , this adds another 10 days of adult and egg developmental time . In Cairns , a cooler winter period will lengthen developmental periods , while dry periods delay hatching . Overall , 10 generations per year is likely to be a reasonable estimate . S2 Fig provides approximations for the areas covered by wMel in different seasons after the releases . For EHW , the area covered in which wMel has at least frequency 0 . 5 is about 1 . 3 km2 in D1 , and this rises to about 2 . 2 km2 in W2 . We can calculate wave speed per generation ( assuming 10 generations a year ) or per day using alternative geometric approximations described in the Methods section: approximation 4 assumes a circular release area , approximation Eq ( 5 ) assumes a rectangle ( for which we approximate parameter y = 2 [i . e . , a release area twice as long as wide] ) , or approximation 6 assumes a rectangle in which spread does not occur ( or is not monitored ) in one direction . ( Note that very little spread occurred to the south at EHW . ) The resulting estimates of wave speed per day are , respectively , cd = 0 . 35 m per day , 0 . 31 m per day , and 0 . 45 m per day . If we assume 10 generations per year ( and so 15 generations separating D1 from W2 ) , the corresponding wave speeds per generation are: c = 12 . 9 , 11 . 2 , and 16 . 6 m/gen . Assuming dispersal parameter σ ≈ 100 m/ ( gen ) 1/2 and unstable equilibrium p^≈0 . 3 ( see Turelli and Barton [31] ) , the cubic diffusion approximation for wave speed ( see Eq 2 ) , c=σ ( ½–p^ ) , predicts roughly 20 m/gen . As discussed in the context of our likelihood analyses below , the discrepancy between the estimated speeds and this analytical prediction can be resolved by assuming longer generations , a higher unstable point , and/or long-tailed dispersal [31] . For PP , the area covered in which wMel has at least frequency 0 . 5 is about 0 . 65 km2 in D1 , and this rises to about 1 . 17 km2 in W2 . Using our geometric Models 4 , 5 and 6 ( with y = 2 , as for EHW ) , the resulting estimates of wave speed per day are , respectively , cd = 0 . 28 m per day , 0 . 25 m per day , and 0 . 37 m per day . If we assume 10 generations per year ( and so 15 generations separating D1 from W2 ) , the corresponding wave speeds per generation are: c = 10 . 4 , 9 . 0 , and 13 . 4 m/gen . The speed estimates for PP are systematically smaller than for EHW . As discussed in the Methods section , both wave speed and wave width ( describing the distance over which infection frequencies change appreciably ) are proportional to average dispersal distances . Thus , slower wave speed is expected if the higher adult densities observed at PP versus EHW translate into a more desirable habitat and consequently smaller average dispersal distances ( lower σ ) . Consistent with this , we find a sharper wave at PP as quantified by smaller average distances between the 0 . 3 and 0 . 8 contours at PP than EHW; these distances average 326 m at EHW and only 252 m at PP . Our likelihood analyses are independent of the graphical summaries produced by Kriging . They rely on an approximate description of the expected shape of local spread and/or collapse ( see Eq 7 in the Methods section ) . We present several successive analyses that summarize the rate and pattern of spatial spread of Wolbachia at EHW and PP . Our summaries focus on 2 statistics: wave width and wave speed . We start by analysing the data averaged over space and time , then present more detailed analyses that document heterogeneous spread . We begin by analysing the data assuming that observed frequencies deviate from deterministic expectations only because of binomial sampling variation . We then use a more complex probability model that accounts for additional sources of heterogeneity . Finally , we explicitly test for directional heterogeneity in rates of spread , as documented visually in Figs 3 and 4 . The details of the likelihood analyses are relegated to S1 Text . Fig 5 illustrates the slow collapse of the wMel introduction at WC . As shown in S2 and S5 Figs , in contrast to the rising infection frequencies outside the release zones at EHW and PP , p initially rises then slowly falls near the WC release . A likelihood analysis of the pooled data , analogous to those presented in Table 1 , S3 Fig and Fig 6 , supports this conclusion . The details of the analysis are given in S6 Table , with the results graphically summarized in Fig 7 . Unlike the steady outward movement of the wave shown at EHW and PP , with the wave widths stabilizing at values near 400 m , Fig 7 shows that the estimated location , r0 , of the “wave” at WC retreats through time , while the wave width , w , steadily increases , corresponding to slow collapse of the wMel introduction . From our likelihood analyses , the wave speed cd is approximately 0 . 5 m per day ( 186 m per year ) at EHW with wave width w about 460 m . In contrast , we find a slower moving and sharper wave at PP with cd approximately 0 . 3 m per day ( 110 m per year ) and wave width w about 380 m . These estimates are broadly consistent with our heuristic approximations ( from Eqs 4–6 ) obtained from the Kriging plots in Figs 3 , 4 and 5 . As demonstrated by Turelli and Barton [31] , even with fast local dynamics and long-tailed dispersal , we can accurately approximate average local dispersal as σ = w/4 m/ ( gen ) 1/2 ( Eq 2 ) . From this we infer σ ≈ 115 m/ ( gen ) 1/2 at EHW; in contrast , we obtain σ ≈ 95 m/ ( gen ) 1/2 at PP . Given that the support intervals for the estimates of w at EHW and PP do not overlap ( Table 1 ) , we expect these results reflect differences in local dispersal . Given that PP has consistently higher population densities , this difference may reflect less dispersal in a habitat where mosquito densities are higher . However , this needs further testing against alternative hypotheses , such as more dispersal barriers surrounding the PP versus the EHW release areas . It is notable that both the EHW and PP estimates of dispersal are consistent with values obtained from release–recapture experiments ( reviewed in [31] ) . If we assume that the wave speed follows the cubic diffusion approximation c=σ ( ½–p^ ) , per generation and that generations are T days long , we can in principle reconcile observed wave speeds with expected wave speeds at each release site by choosing p^ and T appropriately , namely T=σ ( 12−p^ ) /cd , ( 1 ) where σ is the local dispersal estimate and cd is the observed wave speed per day . For instance , if we assume that at both EHW and PP , p^=0 . 3 , the observed and expected wave speeds can be reconciled if we assume that T = 46 days for EHW , whereas T = 63 . 3 days for PP . Given that population densities are higher for PP , increased crowding may indeed produce longer generation times [53] . These times are systematically larger than our conjecture of 10 generations per year , which we supported by an informal data review above . These inferences assume that the cubic-diffusion prediction for wave speed ( c=σ ( ½–p^ ) per generation ) is accurate for these field populations . However , as shown by Turelli and Barton [31] , long-tailed dispersal with fast local frequency dynamics ( as expected with complete cytoplasmic incompatibility , corresponding to sh = 1 in the models of [31] ) , can slow the expected wave speed by 20%–40% below the cubic-diffusion prediction . If the expected wave speed is reduced by 30% , the observed wave speeds match the modified expectations with generation times reduced to 32 . 2 and 44 . 3 days at EHW and PP , respectively . These times are closer to our conjecture of 10 generations per year . In general , there seems to be reasonable quantitative agreement between the slow observed wave speeds and the predictions of simple models using parameter values that are consistent with the poorly known field biology of Ae . aegypti and the deleterious fitness effects of wMel in Ae . aegypti . Despite many caveats , including uncertainty about parameter values and the imprecise meaning of the one-dimensional unstable point p^ for populations with overlapping generations and complex ecology [27] , the observed spread rates at EHW and PP are clearly consistent with approximation Eq ( 1 ) using plausible estimates of dispersal distance , the unstable point , and generation time . In contrast to EHW and PP , the releases at WC did not lead to clear establishment and certainly did not produce spatial spread ( see Figs 5 and 7 ) . Turelli and Barton [31] provide conditions on minimum release areas ( and maximum dispersal distances ) consistent with spatial spread , allowing for long-tailed dispersal and rapid local dynamics . We expect that p^≈0 . 25–0 . 3 and σ ≈ 100 m/gen1/2 . If these parameter estimates are accurate , the release area at WC is likely to be just below the minimum needed to produce successful local establishment and spread ( see Table 2 of [31] ) . Moreover , the fact that the apparent collapse at WC is extremely slow is consistent with the slow dynamics expected near that critical size threshold for wave-establishing releases [31] . Overall , the bistable dynamics of wMel in Ae . aegypti will impose some minimum release size , and only WC is near a plausible minimum . To rigorously test the minimum-release-area predictions of Barton and Turelli [20] and Turelli and Barton [31] , several more replicate releases in small areas would be needed .
Our data demonstrate that wMel can be stably established locally within urban areas surrounded by uninvaded but suitable habitat . Hence , stable population replacement is not limited to small isolated habitats such as those where the initial releases and establishment of wMel in Ae . aegypti took place ( cf . [14] ) . Moreover , the temporal increase in infection frequency within the EHW and PP release zones was comparable to that seen in the isolated areas . In contrast , the smallest release area , WC , did not show stable invasion . This suggests that there is little impediment to the local establishment of Wolbachia in urban areas , provided the releases are conducted over sufficiently large areas ( e . g . , on the order of 0 . 5 km2 when dispersal distances are comparable to those in Cairns [31] ) . These findings highlight the feasibility of patchy releases across large cities , suggesting that area-wide replacement can be produced gradually , with patchy releases complemented by natural local spread . At EHW and PP , the area in which Wolbachia persists at high frequency roughly doubled after 2 years ( Figs 3 and 4 ) . The failure of wMel to establish and spread at WC seems attributable to the small area of the release zone , as the habitat conditions in and around WC are similar to EHW and PP . This is consistent with mathematical predictions concerning the minimum release zone radius , Rcrit [20 , 31] . Based on the wider advancing wave front seen at EHW versus PP , we infer greater average dispersal distance at EHW ( which is likely to provide fewer feeding and breeding opportunities than PP ) . Mosquito dispersal differences probably explain the faster spread observed at EHW versus PP . In contrast , the slow temporal and spatial dynamics of local infection frequency at WC suggests that 0 . 11 km2 , the area of the WC release zone , may be very close to the minimum size needed to initiate spread , at least for the levels of dispersal typical of Cairns . When contrasted against the successful spread at PP , we conclude that the critical release area under Cairns conditions is somewhere between 0 . 11 km2 and 0 . 52 km2 . In tropical regions that support denser Ae . aegypti populations , we expect lower dispersal distances . This would allow successful local establishment using smaller release areas , but spatial spread would also be expected to be even slower than the 100–200 m per year observed at EHW and PP . The heterogeneity in both the speed and patterns of the spatial dynamics at EHW and PP suggests that local environmental factors greatly influence the spread of Wolbachia transinfections ( such as wMel in Ae . aegypti ) that produce significant fitness costs . Spread at each site exhibited strong spatial structure throughout the study , and the structure persisted across the monitoring period . Areas that were easily invaded during the first dry season after the releases ( D1 , see Figs 3 and 4 ) generally stayed invaded in successive seasons , and the autocorrelation among mosquito numbers and infection frequencies increased as the study progressed ( S7 Table ) . The invasion spread well beyond the initial release zones at EHW and PP , and our likelihood analyses ( Fig 6 ) suggest that slow but steady spread would continue in the absence of further releases until significant barriers to dispersal are encountered . Barriers to spread can include both barriers to Ae . aegypti dispersal and variation in Ae . aegypti population density [20] . At PP , the invasion spread south from the release zone immediately but never established to a high frequency south of Mulgrave Road . Nevertheless , infected mosquitoes were caught at low frequencies south of Mulgrave Road from season W1 onwards . These observations are consistent with the demonstration in Trinidad that roads represent partial barriers to Ae . aegypti dispersal [43] . At the very least , such barriers slow wave propagation [20] . It remains unclear whether Mulgrave Road provides a sufficient barrier to stop the wave of Wolbachia , as is the case of the Bruce Highway at Gordonvale . There , Wolbachia have failed to invade an area adjacent to the 2011 release zone for several years , despite persistent migration across the highway [31] . Other evidence from mark-release experiments and genetic studies have pointed to potential barriers ( roads , rivers , forests ) to movement of Ae . aegypti at a local scale [43] . In W2 at EHW , there was an apparent drop in p in the southern half of the release zone . This was unexpected given that in previous seasons traps in this region had recorded Wolbachia frequencies close to fixation . It appears that the drop was due to a sudden increase in uninfected mosquito numbers onsite , which may represent the hatching of dormant uninfected eggs or an early influx of uninfected mosquitoes from an external source at the start of W2 . One possibility is that Wolbachia infected larvae experienced a fitness cost under high-stress conditions prevailing at that time; such costs have been recently documented under stressful conditions that produce a range of adult sizes [19 , 56] similar to those seen under field conditions [54] , even though earlier studies suggested only modest fitness costs associated with wMel [11 , 16] . The openness of the EHW study area may also make it more vulnerable to reinvasion , as immigration into the release zone was possible from 360° of the surrounding area . In contrast , one of the long edges of the PP release zone was bounded by parkland that blocked immigration and may help explain the slow but uninterrupted spread observed there ( S2 Fig ) . Seasonal variations in invasion dynamics are expected when mosquito abundance varies throughout the year . Models of Wolbachia population dynamics show that when mosquito host abundance fluctuates seasonally , p is expected to decline in the low seasons because of slow recruitment and comparatively high mortality among infected imagoes [57] . This was not observed in PP but was in EHW , where during D2 the area covered by the infection shrank and offsite p dropped considerably , only to recover the following wet season . PP may have been shielded from these effects by its apparent abundance of good mosquito habitat , reflected in its high trap yields throughout the study . Very few mosquitoes were caught in EHW during D2 , and the sluggish recruitment there was a likely cause of the retraction of the invasion during that season . Initial onsite infection frequencies at EHW correlated positively with window screens and negatively with habitat quality ( S8 Table ) . This corroborates the findings of Hoffmann et al . [32] that following mass release of infected mosquitoes within an area , Wolbachia frequencies are highest in areas of poor mosquito habitat . However , no relationship was found between trap yields and simple measures of habitat quality ( S2 Text ) . While BG-Sentinel traps can pick up on seasonal changes in mosquito abundance [58 , 59] , they may not be able to give precise estimates of local Ae . aegypti densities at the scale of deployment used in this study . Modelling offsite spread as a function of easily observed habitat variables was inconclusive ( S2 Text ) . No variables were consistently predictive across seasons or sites , and in some cases variables that were expected to encourage spread ( i . e . , areas of low Ae . aegypti density: those with window screens , low-set dwellings , poor habitat quality ) were found to deter it . However , predictions based on variation in population density and uniform dispersal ( e . g . , [20] ) may be confounded by active searching for favourable oviposition sites [60] , if density variation is driven by local habitat quality . The lack of any discernible predictor variables , the strongly heterogeneous spread , and the drop in infection frequencies at the centre of the EHW release zone during W2 suggest that stochastic processes may have played a role in the invasions of EHW and PP . This is surprising considering that the scale of the Cairns invasions was much larger than those thought to be susceptible to stochastic effects associated with very small numbers of infected individuals [61] . Alternatively , a series of highly localised processes may have influenced the heterogeneity of the spread . If this is the case , our BG-Sentinel traps may be too dispersed to pick up on local variability that could inform future releases [62] . Spatial structure in Ae . aegypti populations has been observed at the house scale in Cairns [63]; in ensuing releases of Wolbachia-infected Ae . aegypti , a more clustered placement of traps within and around the advancing wavefront may provide a clearer picture of the processes at work . Despite the heterogeneity , our simple 2-parameter model seems to plausibly account for the slow rates of spread observed at our larger release sites . Bistability for the wMel transinfection , versus the apparent tendency for successful natural Wolbachia infections to spread even when very rare , accounts for the fact that spread in Cairns is orders of magnitude slower than observed Wolbachia spread in natural Drosophila populations [30 , 31] . In summary , we have found rapid local establishment of wMel Wolbachia in the Ae . aegypti populations of urban release areas , with an adjacent suitable habitat available for mosquito dispersal . In the 2 release areas that exceeded the predicted minimum size threshold for local establishment , the infection remained at a moderately high frequency for 2 years . Moreover , wMel spread slowly outward at a rate consistent with theoretical predictions , based on realistic estimates of local dispersal and the position of the unstable equilibrium frequency . While this rate of Wolbachia spread is extremely slow , these findings indicate that large urban areas can be transformed gradually with patchy local Wolbachia releases [31] . Local information about barriers to dispersal can inform the minimum number of releases required , but it remains a challenge to understand the heterogeneity of spatial spread in terms of easily obtained data concerning habitat quality .
Ae . aegypti adults infected with the wMel strain of Wolbachia were released by Eliminate Dengue staff at 3 zones in Cairns , Queensland , from January to April 2013 . Overall , 131 , 420 mosquitoes were released at EHW , 286 , 379 at PP , and 35 , 196 at WC . More mosquitoes were released at PP because of its denser local Ae . aegypti population , and the 3 sites began with comparable proportions of wild and introduced mosquitoes . The largest of the zones , EHW , was also the most open , situated amid residential suburban development with no potential dispersal barriers in its vicinity . PP and WC were in contrast both semiclosed , with each having 1 side of the release zone bounded by parkland so that from the centrum only 268° of the release zone at PP and 254° at WC was connected directly to urbanised Ae . aegypti habitat . Additionally , both PP and WC were near major roads , specifically Mulgrave Road to the southeast and Captain Cook Highway to the northeast ( Fig 1 ) , which could act as barriers to mosquito dispersal . These roads each consisted of 6–10 lanes totalling >50 m throughout . They were flanked by commercial buildings , for the most part , interspersed with apartment complexes . Due in part to an abundance of modern single-storey housing , EHW was known to support a lower density of Ae . aegypti than the other sites . In contrast , PP was adjacent to Cairns’ central business district , and its household size of 2 . 00 per dwelling—smaller than that of either EHW ( 2 . 29 ) or WC ( 2 . 11 ) ( http://profile . id . com . au/cairns/population ) —reflects a larger number of multistorey apartment complexes , fewer bungalows and a higher density of unscreened older houses . At each of the 3 locations , the area enclosed by the release zone was thought to support a higher density of Ae . aegypti than the area surrounding the release zone , where there tended to be a higher density of modern houses . Cairns experiences a tropical monsoon climate , with a wet season running from November to April . The successful establishment and spatial spread of Wolbachia following releases requires that the release area exceed a theoretical minimum , described by a critical radius Rcrit [20 , 31] . While both EHW and PP clearly surpass the minimum , the small area of WC makes establishment there uncertain . Numerical analyses show that Rcrit depends on both the shape of the dispersal function and the average dispersal distance , with small releases more likely to be successful with lower dispersal that is highly leptokurtic ( i . e . , showing both more long-distance and more short-distance dispersal that expected under a Gaussian function [31] ) . Dynamics of establishment and spread likewise depend on the size of the release zone relative to Rcrit , with faster dynamics predicted when the release zone is very large or very small relative to the minimum size . From this , we expect EHW and PP to display relatively rapid spread . In contrast , given that the release area at WC is close to the minimum , the failure or success of establishment is expected to take on the order of 2 years ( about 20 generations ) to ascertain [31] . BG-Sentinel mosquito traps ( Biogents AG , Regensburg , Germany ) were set up within and around each release zone at fixed positions in the yards of consenting householders , covering a distance of 25–530 m ( EHW ) , 35–670 m ( PP ) , and 30–520 m ( WC ) from the release zones in every available direction . The exact number of trapping sites varied over time as traps were moved from households whose residents had either moved or decided to terminate participation in the project . New traps were also added occasionally . By April 2015 , data had been collected from 182 traps at EHW ( 44 onsite traps within the release zone , 138 offsite traps outside the release zone ) , 142 traps at PP ( 42 onsite , 100 offsite ) , and 74 traps at WC ( 20 onsite , 54 offsite ) . BG-Sentinel traps catch mosquito adults by means of a visual lure ( black entry cup ) and suction fan . In Cairns , they capture Ae . aegypti with high specificity and in large numbers [59] . Traps were checked weekly from 7 May 2013 to 30 April 2015 . Traps that failed because of malfunction , invasion by predators ( ants , spiders ) or physical disturbance were scored as null observations for that time point . Adult mosquitoes from each trap were stored in ethanol at –20°C . Traps at WC ceased being checked after 1 April 2015 , as new releases began in the area . Samples were shipped to Monash University , where Wolbachia frequencies in mosquitoes from individual traps were determined by PCR using methods as previously described [14] , with the following modifications . Samples were run through a multiplex qPCR assay with Taqman probes to detect Wolbachia and confirm identification of Ae . aegypti in the same reaction . Samples were extracted in 50-μL squash buffer ( 10 mM Tris pH 8 . 4 , 1 mM EDTA , 50 mM NaCl ) with 1 . 25% Proteinase K with a mini-beadbeater for 1 . 5 minutes and then incubated at 56°C for 5 minutes , incubated at 98°C for an additional 5 minutes , and then kept at 4°C until run . PCR reactions were run in a 10-μl total volume consisting of Lightcycler 480 mastermix , 1 μl of DNA extract , and primers and probes as follows . Species identification was determined with Ae . aegypti ribosomal protein gene RPS17 using Rps17_FW: 5′-TCCGTGGTATCTCCATCAAGCT-3′ , Rps17_RV: 5′-CACTTCCGGCACGTAGTTGTC-3′ , with Rps17_TaqM_Probe: 5′-FAM-CAGGAGGAGGAACGTGAGCGCAG-BHQ1-3′ . Wolbachia infection status was determined with wMel gene WD0513 using TM513_F: 5′-CAAATTGCTCTTGTCCTGTGG-3′ , TM513_R: 5′-GGGTGTTAAGCAGAGTTACGG-3′ , with TM513_TaqM_probe: 5′-LC640-TGAAATGGAAAAATTGGCGAGGTGTAGG-Iowablack-3′ . Analysis was done by absolute quantification and the second derivative method in Roche Lightcycler software . Ae . aegypti abundances are known to vary considerably throughout the year in Cairns [64] . This led us to partition the trap data into seasonal units , reflecting the 6-month wet and dry seasons of Northern Queensland . The first dry season D1 ( May 2013–October 2013 ) began immediately after the releases , followed by the first wet season W1 ( November 2013–April 2014 ) , the second dry season D2 ( May 2014–Oct 2014 ) , and the second wet season W2 ( Nov 2014–Apr 2015 ) . This allowed comparisons across time as well as space . For our graphical analyses , we aggregated data for each trap to give the following: ( 1 ) total mosquito abundance per season , ( 2 ) total number infected with Wolbachia per season , ( 3 ) an average number of mosquitoes observed per week ( total and infected ) , and ( 4 ) a seasonal infection frequency , p . As in Hoffmann et al . [18] , we checked species identity and Wolbachia infection status by PCR . For each mosquito , PCR was performed using 3 primer sets , Aedes universal primers ( mRpS6_F/mRpS6_R ) , Ae . aegypti–specific primers ( aRpS6_F/aRpS6_R ) , and Wolbachia-specific primers ( w1_F/w1_R ) . | Wolbachia are bacteria that live inside insect cells . In insects that act as viral vectors , Wolbachia can suppress virus transmission to new hosts . Wolbachia have been experimentally introduced into Aedes aegypti mosquito populations to reduce the transmission of dengue , Zika , and other arboviruses that cause human disease . Wolbachia invade populations by causing cytoplasmic incompatibility , a phenomenon whereby embryos from crosses between infected males and uninfected females fail to hatch . While Wolbachia have been shown to successfully invade and remain established in isolated Ae . aegypti populations , outward spread from urban release zones has not been previously documented . This is an important step in demonstrating that Wolbachia can be used to combat mosquito-borne infectious disease in cities . Here we describe Wolbachia spread from 2 introduction areas within Cairns in northeastern Australia at a rate of about 100–200 meters per year . Spread occurs only when introduction areas are sufficiently large . The slow rates of observed spread are broadly consistent with mathematical predictions based on estimated Ae . aegypti dispersal distances , Wolbachia dynamics , and effects seen in isolated populations . Spread is uneven and likely depends on local characteristics ( e . g . , barriers ) that affect mosquito density and dispersal . Our data indicate that Wolbachia can be introduced locally in large cities , remain established where released , and slowly spread from release areas . These dynamics indicate that high Wolbachia infection frequencies can be established gradually across large urban areas through local releases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
]
| [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"animals",
"wolbachia",
"seasons",
"mathematics",
"statistics",
"(mathematics)",
"population",
"biology",
"insect",
"vectors",
"bacteria",
"infectious",
"diseases",
"geography",
"aedes",
"aegypti",
"disease",
"vectors",
"insects",
"arthropoda",
"approximation",
"methods",
"population",
"metrics",
"mosquitoes",
"statistical",
"models",
"urban",
"areas",
"earth",
"sciences",
"geographic",
"areas",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"population",
"density",
"physical",
"sciences",
"organisms"
]
| 2017 | Local introduction and heterogeneous spatial spread of dengue-suppressing Wolbachia through an urban population of Aedes aegypti |
Interactions in protein networks may place constraints on protein interface sequences to maintain correct and avoid unwanted interactions . Here we describe a “multi-constraint” protein design protocol to predict sequences optimized for multiple criteria , such as maintaining sets of interactions , and apply it to characterize the mechanism and extent to which 20 multi-specific proteins are constrained by binding to multiple partners . We find that multi-specific binding is accommodated by at least two distinct patterns . In the simplest case , all partners share key interactions , and sequences optimized for binding to either single or multiple partners recover only a subset of native amino acid residues as optimal . More interestingly , for signaling interfaces functioning as network “hubs , ” we identify a different , “multi-faceted” mode , where each binding partner prefers its own subset of wild-type residues within the promiscuous binding site . Here , integration of preferences across all partners results in sequences much more “native-like” than seen in optimization for any single binding partner alone , suggesting these interfaces are substantially optimized for multi-specificity . The two strategies make distinct predictions for interface evolution and design . Shared interfaces may be better small molecule targets , whereas multi-faceted interactions may be more “designable” for altered specificity patterns . The computational methodology presented here is generalizable for examining how naturally occurring protein sequences have been selected to satisfy a variety of positive and negative constraints , as well as for rationally designing proteins to have desired patterns of altered specificity .
Proteins have evolved to operate within the context of crowded cellular milieus and complex functional networks [1] . It is not well understood how and to what extent protein sequences and structures are optimized for multiple and likely interdependent properties such as stability and efficiency of folding , low propensity for aggregation , and functional characteristics . Protein–protein interaction networks may impose particular pressures on amino acid residues in protein interfaces if each protein needs to maintain correct and avoid unwanted interactions . Not only specificity of interactions but also a defined level of promiscuity may be required , as it is known that many proteins use regions of overlapping interface residues to bind several partners at different points in time [2] . Protein design predictions offer great promise to help dissect the structural determinants of the interplay between promiscuity and specificity , as well as to create new molecules that interfere with defined cellular protein–protein interactions with high fidelity and selectivity [3] . Protein design methodologies have generally focused on choosing an amino acid sequence optimal for a specific criterion , such as protein stability or interaction energy with a single binding partner . These computational design techniques have led to several accomplishments , including the pioneering design of a complete protein [4] , of novel protein folds [5 , 6] , the engineering of catalytic activity into an uncatalytic scaffold [7 , 8] , and the redesign of protein–protein interfaces [9–12] . Yet , sequences completely redesigned on a known protein fold often differ substantially from naturally occurring sequences [4] , and the properties of designed proteins can be unusual . Top7 , a computationally designed protein with a fold not previously seen in nature [6] , has a complex folding landscape strikingly different from that of evolved small , single domain proteins [13] . Thus , if we wish to rationally design new proteins that can be expressed and function correctly in a cellular environment and in the context of many possible interaction partners , it is likely that we will need modeling procedures that are able to consider a variety of requirements defining optimal protein “fitness . ” Here we focus on the multiple constraints interaction networks may impose on protein interfaces , both to characterize the evolutionary and biophysical principles shaping these networks , and to develop computational design methods to reengineer them . Previous studies have suggested the importance of negative selection to maintain specificity for a single binding partner [14] . Havranek and Harbury developed a negative design strategy selecting against unwanted partners to predict highly specific coiled-coil interfaces [11] . We extend the idea of incorporating additional selection constraints into computational protein design by examining the inverse problem: how are multiple positive criteria , such as the binding of different partners , accommodated at multi-specific ( e . g . , promiscuous ) protein interfaces ? We perform two computational experiments on 20 multi-specific proteins . First , we optimize each multi-specific interface sequence to maintain interactions with all known structurally characterized partners ( multi-constraint protocol ) . Second , we predict interface sequences optimal for interacting with each partner individually ( single-constraint protocol ) . We hypothesize that , to the extent that a multi-constraint protocol is a good approximation of pressures acting on promiscuous interfaces , predicted sequences should be more “native-like” when all characterized binding partners are included in the optimization procedure than if only the interaction with a single binding partner is considered . Further , we can compare the differences in interface sequences selected by each partner alone ( single-constraint ) and all partners considered together ( multi-constraint ) . If multiple pressures play a role for sequence choices , this comparison should highlight which amino acids are compromises among the various outcomes favored by each binding partner individually . We show that , overall , inclusion of multiple binding partners during optimization returns sequences closer to those found in native promiscuous interfaces . We find native interface residues predicted to be “hotspots” for each partner remain optimal in the context of optimization for single or multiple partners , while other positions may or may not undergo compromises in order to maintain binding of all partners . These trends resulted in the classification of two broad groups of multi-specific interfaces . In the first group , the number of native residues recovered as optimal was similar for optimizations performed over single or multiple partners . Here , key interactions within the interface appeared to be shared , and there was little evidence of compromise in binding preferences among all partners . In contrast , a second group of multi-specific proteins , including “hubs” such as small GTPases , ubiquitin , and actin , appeared to have optimized large fractions of their interfaces for binding multiple partners . In these cases , each partner appears to pick and choose subsets of the interface to make key interactions with , and integrating differences in the binding preferences over all partners often resulted in the native residues being the “optimal compromise” for maintaining binding of all partners . Our method thus both predicts interface sites responsible for multi-specificity and provides an estimate of the magnitude of pressure exerted on sites by each interaction partner . The method we present here can be used as a predictive tool to study how naturally occurring amino acid sequences might have been constrained by any number of positive or negative criteria—including the ability to adopt two different conformations [15]—or as a protein design tool to rationally redesign variants of native proteins to have a desired set of properties matching user-defined constraints .
We set out to address two main questions ( see Figure 1A ) . First , how optimized are native multi-specific interface sequences for binding multiple partners ? It is known that the free energy of binding a single interacting partner is generally not evenly distributed among the native interface residues , but rather some hotspot positions are energetically more important than others [16] . Further , phage display experiments have revealed that substantial sequence plasticity may be tolerated at protein interfaces without significantly destabilizing , and often improving , binding of a single partner [17–19] . Thus , only a subset of a protein sequence may need to be constrained to native in order for a single criterion to be satisfied , while other sequence positions may be less optimized and tolerate a wider set of amino acid types . We hypothesize that the presence of multiple constraints ( e . g . , multiple binding partners ) might substantially restrict interface sequence space such that only native or near native amino acid residues would be tolerated at most sites in multi-specific interfaces . If this is true , sequences that have been computationally designed to optimize binding with all known interaction partners should be more “native-like” than sequences designed to bind each partner independently . Thus , for each promiscuous protein we examine , we compare the sequence predicted to be optimal by our multi-constraint protocol to the wild-type sequence in order to provide an estimate of how extensively each interface is optimized for multi-specificity ( see Test for Optimization in Figure 1A1 ) . Importantly , differences between predicted and wild-type sequences could highlight that evolved sequences are not necessarily optimized for maximal affinity but that other pressures are at play . Second , we ask if each binding partner prefers similar interactions throughout the binding site , or if some partners need to make energetic compromises in order for multi-specificity to be maintained . To address this question , we compare sequences computationally designed to bind only one partner at a time ( without consideration of the other characterized partners ) with the sequence selected as optimal for interacting with all partners ( see Estimate Cost in Figure 1A2 ) . We reasoned that for a given interface position , if an identical amino acid is chosen when each partner is optimized separately as is selected when all partners are included in the optimization process , key interactions at that site might be highly shared among all partners . In contrast , some single-constraint optimizations might choose an amino acid type different from the one selected in the multi-constraint protocol . For such positions , we can use our scoring function to estimate the degree of compromise occurring between differing preferences seen among the interaction partners . We imagined two extreme case scenarios . If all binding partners of a given promiscuous protein prefer similar interface sequences ( “shared” scenario ) , single- and multi-constraint optimizations are expected to give similar results and comparable agreement with wild-type sequences ( termed “native sequence recovery” ) . If only a core set of shared residues is sufficient for binding to all partners , the total native sequence recovery over the entire interface could be low , as the exact amino acid identity of peripheral residues may be less important . Alternatively , each residue in a multi-specific interface could be optimal for only one or few partners ( “multi-faceted” scenario ) . In this case , designed sequences from single-constraint simulations would be expected to resemble the wild-type sequence only for certain positions , and these positions could be different for each partner . The multi-constraint simulations should act to integrate preferences across all partners and would be expected to generate sequences that are more native-like than those resulting from optimization for any single binding partner alone . For this scenario , there could be significant tradeoff between the preferences of differing partners , and amino acid residues within this class of interfaces could be compromises with respect to the amino acid type preferred by some or all partners . However , for each interface position , we hypothesize that there should be an “optimal compromise” that satisfies the constraints imposed by all partners to maintain multi-specific binding . Our computational protocol to test for optimization and compromise in multi-specific interfaces outlined above is illustrated schematically in Figure 1B . To determine whether the shared or multi-faceted strategies are used in naturally occurring promiscuous interfaces , we first compiled a dataset of protein complexes from the PDB ( Protein Data Bank ) ( see Methods ) . Each promiscuous protein along with all its structurally characterized binding partners is listed in Figure 2 . In total , we examined 65 PDB complexes , each of which included one of 20 multi-specific proteins . While this analysis is inherently limited by the set of promiscuous proteins characterized in the PDB and ignores much known information on biological interactions , it has the advantage that we can rely on high-resolution structural information for each of the complexes , and hence are more likely to obtain reliable predictions from protein design methods . Our dataset of 20 promiscuous proteins is nevertheless quite broad and includes all SCOP ( Structural Classification of Proteins ) [20] classes ( except membrane proteins ) as well as representatives from diverse functional families such as signaling proteins ( GTPases , CheY ) , structural proteins ( actin ) , ubiquitin , and several enzymes ( see Figure 2 ) . Further , in order to estimate the connectivity or number of putative protein–protein interactions for each promiscuous protein in our dataset , we performed a BLAST search against sequences within the Database of Interacting Proteins ( DIP , http://dip . doe-mbi . ucla . edu/ [21] ) . Protein–protein interaction graphs for homologs to the multi-specific proteins in our set ( leftmost column of Figure 2; e-value < 1 * 10−9 , see Methods ) suggest that at least half of the proteins we analyze can be classified as “hubs” ( first shell nodes > 5; second shell nodes > 15 ) and that many of these proteins are involved in cellular signaling processes . As we wished to examine promiscuous interface positions believed to be under multiple constraints , only interface positions that had an atom within 4 Å of two or more separate binding partners were considered in our analysis . On average , each characterized binding partner contacted 15 ( ±4 . 5 ) residues in this overlapping set ( see Figure 2 ) . Any conformational changes occurring between the different complexes were taken into account implicitly by using the backbone conformations directly from each complex PDB structure . All computational protein design experiments used RosettaInterface and RosettaDesign , which have previously been used to predict binding energy hotspots in protein–protein complexes and to reengineer specificity in protein interfaces [12 , 22] . The scoring function [6 , 22] is dominated by atomic packing interactions , an orientation-dependent hydrogen potential [23] , and an implicit solvation model [24] . Side chain rotamers were modeled on a fixed backbone , and optimal rotameric conformations were chosen for each complex backbone using a Monte-Carlo simulated annealing protocol . Sequence optimizations used a genetic algorithm [11] , and fitness for binding was evaluated using inter-molecular scores ( see Methods ) . Single-constraint optimizations minimized the binding score for interaction with a single partner while multiple-constraint optimizations minimized the sum of the calculated binding scores over all partners ( see Methods ) . Before discussing results over the entire dataset ( complete data for all promiscuous proteins in our set are available as Tables S1–S20 ) , we consider as a representative example the promiscuous protein Ran with five of its structurally characterized interaction partners ( Figure 3 ) . One multi-constraint and five single-constraint optimizations were performed for the Ran set . The trajectories of the five independent single-constraint optimizations monotonously decrease in score at each generation , and in each case the converged final sequence is predicted to have a binding score better than wild-type ( Figure 3A , crosses at final generation ( right edge of the graph ) ) . Additionally , the sequences selected as optimal in each single-constraint simulation differ significantly from native ( 22%–39% native sequence recovery , plus signs in Figure 3C ) . In contrast , the trajectories of the multi-constraint simulation show correlated changes in binding scores as each sampled sequence is evaluated separately in the context of the five complexes ( Figure 3B ) . Cases where the simulation makes tradeoffs that are more favorable to some partners and less favorable for others can be clearly seen ( arrows in Figure 3B ) . Here , the sum of scores over all complexes decreases with time and the final converged sequence ranks closer to the native score than the sequences selected by the single-constraint optimizations ( compare endpoints of trajectories of Figure 3A and 3B with crosses at the final generation ) . Most notably , the amino acid sequence selected as optimal by the multi-constraint protocol is quite similar to the evolved wild-type sequence ( 67% identical to wild-type , plus signs in Figure 3C ) . In the Ran example , the high native sequence recovery seen in the multi-constraint optimization indicates that a significant fraction of wild-type residues in this promiscuous interface is optimized for multi-specificity by “adding up” information from single-constraint optimizations ( Figure 3C ) . This is consistent with the multi-faceted scenario described above . Further , the multi-constraint trajectories illustrate that there may be tradeoffs in preferences among the binding partners ( Figure 3B , arrows ) , and comparison of sequences selected by the single- and multi-constraint simulations suggest interface positions where the wild-type residue may represent a compromise to allow promiscuity . Figure 4A–4F depicts one such instance where several single-constraint optimizations select residues differing from native , yet the multi-constraint optimization integrates the single partner preferences to recover the wild type-glycine ( single-constraint models shown in Figure 4A–4F are for the interface region around residue 74 , first box in Figure 3C ) . The design simulations predict that three of Ran's binding partners ( Figure 4A , 1A2K . pdb; Figure 4C , 1IBR . pdb; Figure 4D , 1K5D . pdb ) prefer side-chains larger that the wild-type glycine that have additional side-chain hydrogen bonding capability . However , tight steric constraints for binding the remaining two partners ( Figure 4B , 1I2M . pdb , and Figure 4E , 1WA5 . pdb ) necessitate glycine to be the “optimal” compromise for this interface position . Similar instances of compromise at interface positions that are under substantial steric constraint with a subset of the interaction partners are a common pattern in our dataset; many of these cases involve wild-type glycine residues . In contrast to the compromised scenario described above , Figure 4G–4J ( multi-constraint models shown in Figure 4G–4J are for the interface region around residue 76 , second box in Figure 3C ) depict a Ran interface residue that our simulations predict to be highly shared among all partners . Here the wild-type residue , arginine , is correctly recovered by every single-constraint simulation where it mediates an inter-chain hydrogen bonding network . This is the case for all partners except one ( see Figure 4I ) . Here the interchain interactions are formed largely by the aliphatic part of the arginine side chain , and design simulations favor a leucine residue . Hence , for this interface position , where the multiple-constraint simulation also correctly selects the wild-type arginine , there is little indication that recovery of this native amino acid is the result of compromises among the interaction partners . Interestingly , the Ran interaction partners depicted in Figure 4F and 4G form very similar hydrogen bonding interactions with the wild-type arginine , although the partner proteins comprise different fold classes . This behavior of physicochemically similar interactions formed by structurally distinct interfaces has been observed previously [25 , 26] . We next investigated whether the trend of optimization for promiscuity using the multi-faceted scenario we observed for Ran was common in our dataset . In total , 65 separate single-constraint optimizations and 20 multi-constraint optimizations were performed ( Figure 2 ) . Figure 5A shows that , over our entire dataset , sequences predicted as optimal by the multi-constraint protocol are more native-like than the sequences selected in the corresponding single-constraint runs ( compare distance from red squares of black diamonds or of grey circles ) . There was only one instance ( elastase complexed with inhibitors , promiscuous protein set #9 ) where the single-constraint optimization for binding one of the partners outperformed the multi-constraint protocol in native amino acid recovery . Upon closer look at the pattern of interface residues recovered as native in each case , there seem to be two broad groups of multi-specific interfaces represented in the dataset . About half of the proteins comprised group I ( blue shading , Figure 5A ) , for which the improvement in native sequence recovery in multi-constraint optimizations over single-constraint optimizations was small and total native amino acid recovery was low , regardless of interface size . As described for the shared scenario above , the low native sequence recovery could be due to all interaction partners binding via a few key residues , with the residues peripheral to these free to vary in sequence . This behavior is likely for several group I proteins including elastase , ovomucoid inhibitor , and the SH3 domain complexes . These proteins bind their targets within a narrow groove or cavity , and in addition a considerable fraction of interactions may be mediated through backbone contacts [27] . Low native sequence recovery in group I could also be influenced by inclusion of cross-species interactions ( enzyme-inhibitor complexes and interleukin 6 receptor binding to mammalian and viral interleukin ) as well as lack of sufficient constraints to fully specify the wild-type sequence ( see discussion below ) . In contrast , for the other half of the proteins in our dataset ( group II ) , sequence optimization over all characterized binding partners resulted in significant improvements in native sequence recovery compared with optimizations for binding to a single partner ( pink shading , Figure 5A ) . Here , as described for the multi-faceted scenario above , the multi-constraint optimization procedure was able to “add up” differing amino acid preferences among partners . The resulting high recovery of native amino acids indicates that binding interfaces for proteins in this group are optimized for multi-specificity . Additionally , as compared with group I , group II proteins tended to use larger and flatter interfaces to mediate binding , were more likely to show high connectivity in protein–protein interactions networks , and bound interaction partners with a greater number of different fold types ( see Figure 2 ) . Although generalizations of our conclusions are necessarily limited by the restricted size of our dataset of 20 proteins , a “multi-faceted” recognition pattern spread over a large interface may be a common strategy used by highly connected signaling hubs to bind diverse partners . We have shown that for about half the multi-specific proteins in our dataset ( group II ) , the multi-constraint–designed sequences were substantially more native-like than single-constraint sequences ( Figure 5A ) . According to our rationale outlined above , this suggested a significant level of optimization for multi-specificity in these interfaces . However , not all interface positions were predicted to be native-like , and native sequence recovery over the whole interface in multi-constraint simulations varied between 40% and 71% in this group . Non-native amino acids could be chosen by our optimization protocol because they are predicted to be more favorable than the wild-type residue or , alternatively , because a number of different amino acid types are allowed at a certain position without substantial energetic differences . To test whether the non-native interface residues selected by the design simulations were predicted to lead to significant interface stabilization , we compared the binding scores of sequences selected by the single- and multi-constraint protocols with the scores of the wild-type sequences . For both group I and group II , optimization for only a single binding partner always resulted in a favorable decrease in predicted interface binding score ( Figure 5B , grey line ) relative to the score of the wild-type amino acid sequence ( Figure 5B , red line ) . The binding score patterns for multi-constraint optimizations ( Figure 5B , black line ) , however , differed among the two groups: multi-constraint binding scores were often similar to single-constraint scores for group I proteins ( compare black and grey lines , blue shaded box ) , while for group II proteins multi-constraint binding scores were much closer to those calculated for the wild-type sequences ( compare black and red lines , pink shaded box ) . The division of our dataset into two groups suggested by the native sequence recovery results ( Figure 5A ) was thus mirrored in the predicted binding score patterns for wild-type and designed sequences ( Figure 5B ) . Our simulations suggest that for group I proteins , where sequences and binding scores for single- and multi-constraint optimizations were similar , there might be non-native amino acids which could improve the promiscuous compromise and at the same time strengthen each interaction with each binding partner alone . In contrast , non-native amino acids selected for group II proteins in multi-constraint simulations are predicted to offer little improvement over the binding scores of the original wild-type sequences; this confirms our notion of high levels of optimization for multi-specificity in this group . Interestingly , while our simulations sought solely to maximize binding affinity for each partner , and did not explicitly consider either the relative binding affinities among partners or that naturally occurring interfaces often need to be transient , incorporation of multiple constraints alone was often sufficient to predict sequences with binding scores near or identical to that calculated for native sequences . We next investigated , on a per-residue basis , at which interface positions our optimization protocols predicted native residues to be suboptimal . Experimental analysis of residues critical for maintaining binding with respect to a single interaction partner have shown that often only a subset of the interface comprises key hotspot residues optimized for binding [16 , 17] and that other non-hotspot positions may show a high degree of plasticity [19] . We thus wished to examine how often native residues were being recovered as optimal by our single- and multi-constraint simulations at positions calculated to be energetically important hotspots . For each binding partner , we calculated the per-residue score of the native residue at every interface position , and labeled sites with a native per-residue score of less than −2 as a predicted hotspot . Next we calculated for each position the difference in score between the residue selected by each of our protocols and the score of the native residue ( see Test for Optimization in Figure 1A1 ) . We reasoned that small score differences ( <1 score units; scores are parameterized to approximate kcal/mol [22] ) should reflect that a given optimization protocol recovered the native ( or energetically similar to native ) residue during optimization , and large score differences ( >1 score units ) should indicate the extent to which a non-native residue is predicted to improve binding affinity over native . At hotspot positions , whether optimizations were performed with respect to single or multiple partners , native ( or energetically equivalent ) residues were recovered for each partner with high fidelity ( Figure 6A and 6B , wheat bars , 244/303 and 272/303 for single- and multi-constraint optimizations , respectively ) . This inability to predict non-native residues scoring better than native at hotspot positions was seen for proteins in both group I and group II ( see Figure S1 ) . In contrast , at non-hotspot positions , the native residue was predicted to be suboptimal ( yellow , orange , red bars ) with respect to binding a single partner in approximately half of all instances ( Figure 6A , “all other residues” , 350/682 ) . This is in agreement with experimental phage display data showing the native residue to often be suboptimal for binding at non-hotspot positions [19] . When considered in the context of binding multiple partners , however , these same non-hotspot sites often are now predicted to be suboptimal in only 14% of all instances ( “All other residues” in Figure 6B; yellow , orange , red bars 167/682 ) . Thus , we find that the need to maintain multi-specificity imposes constraints primarily on non-hotspot residues . This results in native residues being recovered more often at such sites as they become the “optimal compromise” for binding of multiple partners . This trend for increased recovery of native residues at non-hotspot positions during multi-constraint simulations was much stronger for proteins in group II than in group I ( see Figure S1 ) . Finally , we wished to estimate the extent of compromise each multi-specific protein in our dataset made in order to maintain binding to all its partners compared with the “ideal” interaction it could have if only a single partner was considered ( see the section Rationale , and Estimate of Cost in Figure 1A2 ) . For each site within an interface , each partner was assigned a “compromise value” ( ranging from 0 to 2 ) . Compromise values were defined as the per-residue difference in score of the amino acid selected when each partner was optimized alone ( single-constraint ) and the residue selected at the same site when all partners were included in the optimization protocol ( multi-constraint ) . The interface site itself was then assigned the largest compromise value seen among all binding partners . For each position in the interface , this number should provide a rough estimate of the maximal amount of tradeoff paid by any partner due to the necessity of other partners binding via the same site ( see Methods and Figure 1A2 ) . Small compromise values ( 0–1 score units ) should indicate that all binding partners prefer the same ( or similar ) residue type as optimal , regardless of the presence or absence of other binding partners . Larger values ( >1 score units ) suggest that for at least one partner , a non-native amino acid is predicted to make more favorable interactions than the wild-type , but may not be tolerated when preferences of all additional binding partners are considered . Figure 7A shows , over our entire dataset , the percentage of sites within each protein interface calculated to have a compromise score between 0 and 0 . 5 . These positions are predicted to be essentially shared , in that no partner considered would have to give up potential gain so that other partners could fulfill their optimal interactions . While we observed a continuum ranging from interfaces calculated to have few completely shared interactions ( all GTPases , actin , ubiquitin ) to those for which the majority of interactions were shared ( inhibitor complexes , SH3 domain ) , this analysis largely confirmed our earlier grouping of the multi-specific proteins within our dataset ( Figure 7A , pink and blue boxes ) . A few group I proteins showed levels of compromise similar to that seen in group II . Interestingly , at least two of these proteins , importin beta ( set #2 ) and cheY ( set #4 ) , were also calculated to be protein interaction “hubs” in our earlier analysis ( see Figure 2 ) . These proteins may thus also employ a “multi-faceted” binding strategy , and the low native sequence recovery seen with the multi-constraint protocol is likely due to our computational prediction being under-determined ( since we lack structural information for a more complete set of binding partners ) . Likewise , we note that among the group II proteins , for IGG1-FC ( set #15 ) many interactions were predicted to be shared by all binding partners , a result that is consistent with an earlier structural analysis of these proteins by Delano et al . [25] . To illustrate the three-dimensional distribution of predicted compromises in multi-specific interfaces , we generated color-coded mappings of compromise scores . Figure 7 shows representative maps for three promiscuous protein interfaces calculated to display high ( Figure 7B , Ran ) , medium ( Figure 7C , CheY ) , and low ( Figure 7D , Ovomucoid Inhibitor ) overall compromise ( maps for the entire dataset are given in Figure S2 ) . Throughout our dataset , higher compromise scores often occurred along the periphery of a binding site , while highly shared residues tended to be more centrally located . While further analysis is needed , this could indicate strong , shared interactions with core hotspots may be necessary for each partner to bind , but that it is along the rim of the overlapping interface site where compromises among the binding partners have to be integrated in order to maintain multi-specificity . This is reminiscent of the idea that hotspot residues necessary for binding often occur in interface cores sequestered from solvent , whereas other non-hotspot parts of the interface , possibly around the rim , account for recognition [16] . Our energetic analysis suggests that many positions within naturally occurring multi-specific interfaces have been optimized for binding to multiple partners , while some native amino acids are predicted to be sub-optimal in the context of single or even multiple partners . Over the entire dataset , the multi-constraint protocol recovered the native interface residue as optimal for just under half ( 161/338 ) of all interface residues examined . Ultimately , experimental data are needed to verify whether choices of non-native amino acids by our multi-constraint optimization protocol are incorrect predictions by our energy function , or whether the predicted choice would indeed strengthen binding for all partners . In general , experimental data validating binding affinities of sequences predicted by our single- and multi-constraint simulations with all interaction partners were not available . However , we did observe one notable case where we could compare one of our predictions of an improved interface with direct experimental data . This occurred for the third domain of turkey ovomucoid inhibitor ( set #3 ) at the key P1 position at which the inhibitor ( or natural substrate ) residue extends into a deep binding pocket . The predicted per-residue binding score at this site suggested that the wild-type residue was a hotspot crucial for maintaining binding with all partners , yet our multi-constraint protocol predicted a non-native amino acid residue to be significantly preferred over native by all partners . As discussed above , prediction of a native hotspot residue to be suboptimal was an infrequent occurrence throughout our dataset ( see hotspots in Figure 6; yellow , red , orange bars ) . Binding affinities for ovomucoid inhibitor mutants containing all 20 amino acids at the P1 position have been experimentally characterized for six different serine proteases [28] . This allowed us to compare the experimental preferences at the P1 position for the two serine proteases complexes in our dataset ( chymotrypsin and SGPB , see Figure 2 ) with the computational predictions . The residue chosen at this site by the multi-constraint protocol , a phenylalanine , was ranked experimentally as the third and fourth most favorable residue for chymotrypsin and SGPB , respectively . There was no amino acid choice more favorable in common for both proteins and the native lysine residue was ranked eleventh and eighth , respectively . We note that while the multi-constraint protocol correctly selected the optimal choice for binding the two characterized binding partners in our dataset , other amino acid types may be optimal for selectively binding different combinations of the six serine proteases studied . Interestingly , the P1 residue of ovomucoid inhibitor is known to vary significantly in nature , with eight differing amino acid types occurring at this position in the 153 avian species analyzed [28] .
Our study uses a protein design method that can in principle be applied to computationally select amino acid sequences under any set of positive and negative constraints that can be defined by a fitness function . Here we have made comparisons between single- and multi-constraint predicted and naturally occurring sequences to quantify optimization and compromise in multi-specific interfaces . Our analysis indicates that first , the protocol presented here is able to detect optimization for multi-specificity in promiscuous interfaces , as sequences and binding scores from multi-constraint simulations are closer to native than those obtained in single-constraint optimizations . Second , we identify two distinct mechanisms for achieving multi-specificity: ( 1 ) shared or low compromise interfaces where a small subset of interface residues have been optimized such that all binding partners utilize this set as hotspots and ( 2 ) multi-faceted or intermediate compromise interfaces where a far larger percentage of the interface has been optimized for multi-specific binding and each partner picks and chooses a subset of interface residue interaction with which to make key interactions . Signaling proteins with large , flat interfaces fall clearly within the “multi-faceted” group II , while enzymes , motif recognition domains , and receptors with smaller , narrower binding interfaces are often found within the shared group I . We speculate that the “multi-faceted” mode might have an evolutionary advantage for signaling interfaces , as here the chance that single mutations will deleteriously affect all binding interactions is reduced . On the other hand , a single mutation may substantially alter the pattern of interaction partners by now favoring certain interactions over others . In this way , multi-faceted interfaces may be more “evolvable” for new sets of interactions . It is interesting to note that the ability to a priori predict binding sites from surface sequence conservation or surface cavity size has been shown to be easiest for proteins similar to those classified as “shared binding” by our methodology [29] . This is consistent with our observations , as in these cases there should be shared evolutionary pressure for conservation of key surface residues by all partners . In contrast , for proteins predicted to display some degree of compromise among the differing binding preferences of their multiple partners , evolutionary pressures could differ depending on which subset of binding partners is most strongly selected for over time . Further , allowing each partner to pick and choose its own subset of interface amino acids for key interactions , as in the multi-faceted case , could necessitate large , easily accessible ( i . e . , flat ) binding surfaces with a certain degree of conformational flexibility; this mechanism could hence partly explain why flat surfaces and conformational variability are frequently seen in multi-specific signaling proteins such as G-proteins [30] . We hypothesize that there should be significant differences in the ease with which binding specificities among partners could be rationally modified and/or small molecule inhibitors could be designed for proteins exhibiting the two modes of multi-specificity described here . The patterns of varying amino acid preferences among different binding partners revealed by comparing the single- and multi-constraint protocols suggest mutations at specific interface positions that could rationally change the specificity or promiscuity seen among binding partners . However , these same factors might make drug or small molecule design toward “multi-faceted” interfaces more difficult . For the group II interfaces in our set , the different partners display varying interface residue preferences ( see Figure 3C ) , and there may be a substantial number of constrained residues in each binding interface ( see Figure 7 ) . Hence , proteins using this mode of interaction may have fairly distributed hotspots that are difficult to interfere with by a small molecule targeted to a single region . A caveat of our study is that first , generalizations may be somewhat limited because of the restricted size of our dataset of high-resolution structures . Second , the results presented here are necessarily dependent on the quality of the scoring function used for optimizations . However , improvement in native recovery seen in multi-constraint simulations could not be directly due to energy function biases , as the same scoring function was used for all simulations . The ability of the Rosetta scoring function to predict energetically important residues has been analyzed previously [22] . We note that amino acid types for which our simulations consistently select the native residue as the best ( optimal ) choice for binding multiple partners include tryptophan , tyrosine , and arginine ( Figure S3; the predicted amino acid frequencies for W , Y , and R closely match the native distribution ) , amino acid types which have previously been shown to be energetically important in binding interfaces [16 , 31 , 32] . Interestingly , where allowed by steric constraints ( for example at the interface periphery ) , we observed an increased selection in our simulations of larger amino acids such as tryptophan , arginine , and histidine , and against smaller amino acids such as alanine , threonine , and valine ( Tables S1–S20 and Figure S3 ) . While this could be due to approximations in our scoring function , an alternative explanation could be that these non-native sequences would , at least in some cases , truly bind more strongly . An overrepresentation of large hydrophobic residues may have been selected against in nature to maintain protein solubility in the absence of binding partners . In addition , while our computational protocol optimizes binding score , naturally occurring transient interfaces may not necessarily have evolved for strong binding . The complexes between small GTPases and their exchange factors ( GEFs ) may be examples of interactions that need to be transient to fulfill their cellular function: in the case of the ARF1-Sec7 interaction , the fungal metabolite Brefeldin A inhibits signaling by stabilizing the complex [33] . It may also be a general trend that multi-specificity must come at a cost of affinity [34] . Additional constraints not explicitly considered in our current protocol , such as selection at the level of on or off rates for complex formation could also account for differences in native and computationally selected sequences . Lastly , we note that while the analysis presented here has focused on the ability of our simulations to identify the wild-type amino acid , strict conservation of a single native amino acid over evolutionary time is rare , and the tolerance for substitution to differing amino acid types can vary between sites in an interface [29] . For example , for the multi-specific protein Ras we found two instances ( Table S12 , positions 32Y and 67M ) where we predicted the interface positions to be energetically important but failed to correctly recover the native amino acid . In both cases , the non-native amino acids selected by our multi-constraint simulations were among the evolutionarily tolerated set seen in a multiple-sequence alignment ( unpublished data , generated as described in Methods ) . A clear extension of our method is thus not only to predict optimal but also a set of tolerated amino acid sequences for a given set of constraints ( ELH and TK , unpublished data ) . While we have applied the multi-constraint design protocol described in this work to examine whether and how promiscuous proteins are optimized for binding multiple partners , the methodology presented here is general and can be extended to analyze how any number of enumerable constraints ( both positive and negative ) affects sequence selection . A logical related analysis would be to characterize the sequence determinants of conformational flexibility where the input constraints would be stability for two or more different conformations . Further , the multi-constraint protocol introduced here is not only predictive of naturally occurring amino acid sequences , but also allows for rational redesign of proteins with altered binding properties which could be instrumental toward understand the role of specificity in protein interaction networks as well as in the engineering of biosensors and new cellular pathways .
Each domain of every protein–protein interface listed in PIBASE ( http://alto . compbio . ucsf . edu/pibase/ [35] ) was classified using the standard SCOP domain definition . SCOP domains were clustered at 90% sequence identity . Clusters containing only intra-protein domain interactions ( only one chain in the PDB file ) were removed , and clusters with duplicates were merged , leaving 168 clusters . Additional filtering via PDB header descriptions to remove multi-subunit , viral coat , and immunoglobulins/MHC proteins resulted in approximately 50 clusters . All clusters containing multiple structures of the same promiscuous protein interacting with differing binding partners using an overlapping binding site ( by visual inspection ) were selected for the dataset of multi-specific proteins . Lower resolution structures of redundant protein–protein complexes were discarded , as well as all structures ( except 1FXT ) determined by NMR . PDB codes of the resulting 20 clusters are given in Figure 2 . All simulations were performed using the RosettaInterface and RosettaDesign methodologies as outlined in [6 , 22] and described below . The Rosetta scoring function is dominated by attractive and repulsive Lennard-Jones interactions , an orientation-dependent hydrogen bonding term [23] , and an implicit solvation model [24] . Side chains from a rotamer library including the native amino acid PDB conformation and with additional rotamers around the χ1 and χ2 angles [4] were sampled on a fixed backbone using a Monte-Carlo simulated annealing optimization protocol . All water molecules , heteroatoms , and hydrogens present in the original PDB were removed , and hydrogen atoms were added as previously described [23] . An initial round of side-chain Monte-Carlo minimization was then performed using the Rosetta scoring function , keeping all amino acid identities and backbone coordinates fixed , while selecting for the optimal rotamer at each side-chain position from the rotamer set as described above . After this initial minimization , all backbone and side-chain positions not determined to be in the shared interface were kept fixed for all subsequent steps . Amino acid positions on each promiscuous protein were considered for single- and multi-constraint design simulations only if any atom of two or more known binding partners was located within 4 Å of any atom of the side chain of interest . For promiscuous proteins with five or more characterized binding partners , only interface positions with an atom within 4 Å of three or more partners were considered . Each single- or multi-constraint optimization allowed all amino acids ( except cysteine ) to be substituted at each position examined . Positions for which the native residue was a cysteine were disregarded . For all simulations , a genetic algorithm was used to generate and propagate putative sequences based on inter-molecular scores , and optimal rotamers for each sequence were chosen separately with consideration of both inter- and intra-molecular interactions by simulated annealing Metropolis Monte Carlo for each fixed backbone as taken from the PDB . This ensured that in the multi-constraint protocol rotameric conformations could differ among binding partners even as identical interface amino acids were scored for each . Simulations were started with an initial random population of 2 , 000 sequences , and the genetic algorithm was allowed to propagate for 100–200 generations . For single-constraint simulations , fitness was defined to be the inter-molecular score for a single complex while for multi-constraint simulations the fitness was a linear sum of the inter-molecular scores of a given amino acid sequence calculated across all characterized binding partners . For all calculations , the weights ( wi ) were set uniformly to 1 . For single-constraint simulations , the sequence that scored optimal with respect to a single complex independently was advanced to the next generation while the multi-constraint protocol advanced the sequence for which the fitness as defined above was minimized . Uniform crossover was used to generate the remaining sequences of the population for the following generation . Random mutation of any given interface sequence was allowed for each generation with a probability of 20% at any given interface position . Simulations converged ( dependent on the size of the shared interface ) on average within 50–130 generations ( see Figure 3A ) . Over the 20 multi-specific proteins in our dataset , 338 interface residues met the criteria for design . Consideration of each interface position in the context of the 65 characterized binding partners resulted in 1 , 199 individual interactions . For each individual interaction , a per-residue inter-chain score was calculated by summing , for any given residue on chain i , pair-wise contributions to the score from all residues on chain j ≠ i . An interface residue was classified as a hotspot for all binding partners for which the per-residue inter-chain score of the original native amino acid in the wild-type complex was calculated to be less than −2 ( see pink shading , Tables S1–S20 ) . Estimates in predicted per-residue improvements ( Figure 6 ) in binding affinity were made by calculating , for each binding partner , the difference in per-residue score of the amino acid chosen by single- or multi-constraint simulations ( Figure 6A and 6B , respectively ) from native . Positions for which the per-residue score for the native amino acid , as well as the amino acid chosen in single- and multi-constraint simulations was zero , were eliminated from the analysis . These 214 positions represented cases where one binding partner did not interact with an interface residue in contact with other partners in our dataset ( see grey shading , Tables S1–S20 ) . Estimates of per-residue constraint ( Figure 7 ) were made by calculating , for each binding partner , the difference in per-residue scores for the amino acid type/rotamer chosen in the single-constraint optimization from the respective score for the amino acid type/rotamer selected by the multi-constraint protocol . The largest magnitude of difference seen among all partners was the constraint value assigned . For simulations that did not recover the native amino acid type , constraint scores between sequences selected using single- and multi-constraint optimization were also calculated and assigned to the native amino acid type . The complete sequence , as taken from the pdb files , of each promiscuous protein in our dataset was searched against all sequences contained within DIP ( http://dip . doe-mbi . ucla . edu/ [21] ) . Hits were considered as significant if they had an e-value of less than 1*e−9 . Protein–protein interaction graphs ( Figure 2 ) were shown for sequences predicted to be homologous to Saccharomyces cerevisiae whenever possible . The DIP identification number , organism , e-value , and assigned DIP protein name for the interaction graph shown in Figure 2 are as given in Table S21 . A multiple-sequence alignment ( MSA ) and evolutionary rates for Ras were calculated using the automated Web server http://consurf-hssp . tau . ac . il for the Consurf-HSSP database [29] using the PDB ID code 1WQ1 . Evolutionary conservation scores ( 1–10 , 10 most conserved ) were 9 and 8 for 32Y and 67M , respectively . 90% ( 186/206 ) of sequences within the multiple-sequence alignment for the native position 32Y contained either a Y or an H , while 89% ( 184/206 ) of sequences at the native position 67M contained H , I , L , M , Q , or V . Multi-constraint simulations selected 32H and 67H as optimal , respectively . | Computational methods have recently led to remarkable successes in the design of molecules with novel functions . These approaches offer great promise for creating highly selective molecules to accurately control biological processes . However , to reach these goals modeling procedures are needed that are able to define the optimal “fitness” of a protein to function correctly within complex biological networks and in the context of many possible interaction partners . To make progress toward these goals , we describe a computational design procedure that predicts protein sequences optimized to bind not only to a single protein but also to a set of target interaction partners . Application of the method to characterize “hub” proteins in cellular interaction networks gives insights into the mechanisms nature has used to tune protein surfaces to recognize multiple correct partner proteins . Our study also provides a starting point to engineer designer molecules that could modulate or replace naturally occurring protein interaction networks to combat misregulation in disease or to build new sets of protein interactions for synthetic biology . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
]
| [
"biophysics",
"none",
"computational",
"biology"
]
| 2007 | Design of Multi-Specificity in Protein Interfaces |
Q fever is a common cause of febrile illness and community-acquired pneumonia in resource-limited settings . Coxiella burnetii , the causative pathogen , is transmitted among varied host species , but the epidemiology of the organism in Africa is poorly understood . We conducted a systematic review of C . burnetii epidemiology in Africa from a “One Health” perspective to synthesize the published data and identify knowledge gaps . We searched nine databases to identify articles relevant to four key aspects of C . burnetii epidemiology in human and animal populations in Africa: infection prevalence; disease incidence; transmission risk factors; and infection control efforts . We identified 929 unique articles , 100 of which remained after full-text review . Of these , 41 articles describing 51 studies qualified for data extraction . Animal seroprevalence studies revealed infection by C . burnetii ( ≤13% ) among cattle except for studies in Western and Middle Africa ( 18–55% ) . Small ruminant seroprevalence ranged from 11–33% . Human seroprevalence was <8% with the exception of studies among children and in Egypt ( 10–32% ) . Close contact with camels and rural residence were associated with increased seropositivity among humans . C . burnetii infection has been associated with livestock abortion . In human cohort studies , Q fever accounted for 2–9% of febrile illness hospitalizations and 1–3% of infective endocarditis cases . We found no studies of disease incidence estimates or disease control efforts . C . burnetii infection is detected in humans and in a wide range of animal species across Africa , but seroprevalence varies widely by species and location . Risk factors underlying this variability are poorly understood as is the role of C . burnetii in livestock abortion . Q fever consistently accounts for a notable proportion of undifferentiated human febrile illness and infective endocarditis in cohort studies , but incidence estimates are lacking . C . burnetii presents a real yet underappreciated threat to human and animal health throughout Africa .
Coxiella burnetii , a zoonotic bacterial pathogen found worldwide except in New Zealand , is transmitted to humans through direct contact with milk , urine , feces , or semen from infected animals as well as inhalation of aerosolized particles from animal placentas , parturient fluids , aborted fetuses , and environmental dust [1] . While infection by C . burnetii in humans can be asymptomatic , symptomatic infection , known as Q fever , can present as an acute undifferentiated febrile illness with the possibility of focal manifestations , such as hepatitis and pneumonia . Acute disease can progress to chronic forms , such as endocarditis , in 0 . 5–2 . 0% of cases [2] , [3] , typically in individuals predisposed by valvular heart disease or immunodeficiency [4] . Q fever is also one of the infectious diseases that has been linked to chronic fatigue syndrome [5] . Infection by C . burnetii has been demonstrated in many animal species , but the principle reservoirs are thought to be sheep , goats , and cattle . In these livestock species infection is often asymptomatic but can cause abortion and reduce reproductive efficiency [1] , [6] . Q fever has recently gained renewed attention after the largest-ever recorded outbreak which involved over 3 , 500 human cases in the Netherlands in 2007–2009 [7] . Recent studies in resource-limited settings have demonstrated C . burnetii as a common cause of febrile illness and community-acquired pneumonia [8]–[10] . Fever etiology research among hospitalized patients in northern Tanzania found Q fever was a more common cause of severe febrile illness than malaria [11] , [12] . As control efforts have led to consistent decreases in malaria incidence throughout sub-Saharan Africa [13]–[17] the diagnosis , treatment , and control of non-malaria febrile illnesses , such as Q fever , are emerging as new public health priorities [12] . In addition to being sources for disease transmission to humans , C . burnetii infection in animals can decrease livestock productivity which can have socioeconomic and indirect health effects on humans , especially among livestock-keeping populations in resource-limited settings [18] . In light of recent findings establishing Q fever as an important cause of severe febrile illness in northern Tanzania [11] , [12] and growing awareness of the potential economic impact of infection in animals , we systematically reviewed the existing literature on the epidemiology of C . burnetii infection among humans and animals in Africa . This survey aimed to consolidate knowledge and identify future research priorities for the following topics: the prevalence of C . burnetii infection in human and animal populations , including surveys of sera or shedding in body fluids; the incidence of disease due to C . burnetii in human and animal populations; risk factors for seropositivity or disease; and infection control efforts undertaken in Africa .
Nine databases were searched with the search string described in Figure 1 including all countries in the 5 United Nations ( UN ) sub-regions of Africa [19] . These search terms were adapted to the particular language of each database , and for those databases that did not allow the combination of Boolean operators , ( q fever ) OR ( Coxiella burnetii ) was used . Two of the databases , CABI and EBSCO Global Health , were searched with the intention to include grey literature . Citations for all years and in all languages were compiled and de-duplicated using EndNote ( Thomson Reuters , New York , USA ) . Abstracts were independently reviewed by two investigators ( SV and MPR ) using combined language competency in English , French , Spanish , and Portuguese or Google Translate ( Google , Mountain View , CA , USA ) and included for full-text review upon meeting predetermined criteria ( Figure 2 ) . Excluded abstracts described studies conducted outside Africa , basic science or immunology experiments , incorrect pathogens , reviews/editorials , case reports or case series , Q fever among returning travelers , periodical lay media content , diagnostic or therapeutic studies without epidemiologic data , theoretical epidemiology , duplicate data published elsewhere , textbooks/manuals , microbiologic studies without epidemiologic data , or arthropod sampling . The same criteria were applied during full-text review and to all grey literature containing sufficient information for adjudication . Cases of disagreement between the two investigators were resolved through independent review by a third investigator ( JEBH ) . Prevalence studies that presented evidence of current or prior infection with C . burnetii in humans and animals were included . We considered the following serologic tests and minimum antibody titer cut-offs for C . burnetii phase I and/or phase II antigen as acceptable evidence of infection in humans and animals based on expert consensus: complement fixation ( CF ) >1∶10 for animals [20] and ≥1∶4 for humans , microscopic agglutination test ( MAT ) ≥1∶4 for humans and animals , indirect fluorescent antibody ( IFA ) ≥1∶25 for animals and ≥1∶40 for humans , and ELISA validated against one of the above methods [20]–[23] . Capillary agglutination tests ( CAT ) were also accepted based on demonstrated correspondence to CF titers [23] , [24] . The terms seropositive or seropositivity are used throughout to describe serologic reactions that met these titer cut-offs . For studies of pathogen shedding , confirmation of C . burnetii by nucleic acid detection , culture , or rodent inoculation was accepted . Studies of the prevalence of C . burnetii infection in humans and animals were classified by extent of study population characterization and sampling strategies , which were categorized as random ( e . g . , proportional , simple cluster , or simple random ) or non-random . Prevalence studies that met these diagnostic criteria and used randomized sampling strategies qualified for data extraction . Prevalence studies with appropriate diagnostic criteria that used non-random sampling were included but did not undergo further data extraction . Studies reporting disease in animals due to C . burnetii were included if the reported cases met the World Organization for Animal Health ( OIE ) case definition for Q fever: abortion and/or stillbirth plus confirmed presence of the bacterial agent , accomplished by 1 ) isolation in culture; or 2 ) polymerase chain reaction ( PCR ) , in situ hybridization , or immunohistochemistry of birth products or of associated vaginal discharge [20] . Evidence of C . burnetii on placental smears with stains deemed appropriate by OIE ( Stamp , Ziehl-Neelsen , Gimenez , Giemsa or modified Koster ) were considered presumptive for disease and included [20] . We also included seroprevalence studies of animals with a history of abortion , as these data , although not of confirmed cases , could yield information about potential associations between C . burnetii exposure and prevalence of animal abortion . For studies of disease in humans , acute Q fever was defined according to the US Centers for Disease Control and Prevention ( CDC ) case definition: a compatible fever syndrome plus four-fold rise in antibody titers to Phase II antigen or detection in clinical specimens by PCR , immunohistochemistry ( IHC ) , or culture [25] . Phase II antigen IFA antibody titer levels for IgG≥1∶200 and IgM≥1∶50 were included as cases based on the high positive predictive value of such results [26] . Studies reporting chronic disease in humans due to Q fever were included if the reported cases met the CDC case definition for confirmed chronic Q fever: culture-negative endocarditis or infected vascular aneurysm , chronic hepatitis , osteomyelitis , osteoarthritis , or pneumonitis with no other etiology plus IFA IgG antibody to C . burnetii phase I antigen ≥1∶800 or detection in clinical specimens by PCR , IHC , or cell-culture [25] . Risk factor studies were evaluated using the same criteria for sampling design and case definitions that were applied to surveys of prevalence or disease , respectively . Risk factor analyses in prevalence studies were excluded if the prevalence study used a non-random sampling strategy . Studies describing control efforts must have presented original data demonstrating the outcome of an intervention to decrease infection or disease incidence in human and/or animal populations . For all qualifying studies , extracted data included study country , city or region , species , population census data when given , sample size , year of study , and diagnostic test as well as the number of seropositive or disease cases , risk factors , or control effort data where applicable . In the case of incomplete data or unclear methods , authors were contacted for further clarification when possible . Descriptive analyses of the extracted data were conducted . No quantitative meta-analysis was undertaken .
Two surveys employed a systematic sampling strategy to assess seroprevalence among linked human and animal populations . The first , from three governorates surrounding Cairo , Egypt , reported C . burnetii seropositivity in 13% of cattle , 23% of goats , 33% of sheep , 0% of buffalo , and 16% of humans in close contact with these animals [27] . The second survey , undertaken in Chad , found 80% of camels , 4% of cattle , 13% of goats , 11% of sheep , and 1% of humans in close contact were seropositive [28] . All other seroprevalence studies sampled only humans or only animal species ( Table 1 ) . Surveys of cattle demonstrated seroprevalence ranging from 4% in Dakar , Senegal [29] , to 55% around the city of Zaria , Nigeria [30] , [31] . Other studies reporting cattle seroprevalence within this range were conducted in coastal Ghana ( 18% ) [32] , Cameroon's Adamawa Region ( 32% ) [33] , southern Chad ( 7% ) [34] , and South Africa's Transvaal Province ( 8% ) [35] . Goat seropositivity ranged from 13% in Chad [28] to 23% in Egypt [27] and 24% in 8 Sudanese states [36] . Surveys of sheep revealed seroprevalences that ranged from 11% in Chad [28] to 33% in Egypt [27] . In Upper Egypt , 23% of dog sera samples indicated prior C . burnetii infection [37] . In studies of pathogen shedding in bovine milk , C . burnetii nucleic acid was detected in 22% of raw milk samples in Upper Egypt [38] . In Zaria , Nigeria , C . burnetii shedding among individual cows was reported in 63% of milk samples from extensively managed cattle and 43% of samples from semi-intensively managed cattle [30] , whereas the prevalence was 26% and 22% , respectively , at the same location one year later [31] . No studies of shedding in fluids other than milk in asymptomatic humans or animals were found by our search , and no human milk shedding studies qualified for data extraction . Seroprevalence in humans ranged from 1% in Chad [28] to 32% in a Nile Delta village in Egypt [39] . In Niamey , Niger , 10% of children ages 1 month-5 years were seropositive [40] , and in Ghana's rural Ashanti Region , 17% of two-year-olds were seropositive [41] . Other surveys reported human seroprevalence at 5% in rural western Côte d'Ivoire [42] , 8% among nomads sampled in rural northern Burkina Faso [43] , and 5% of pregnant women attending an antenatal clinic in Dar es Salaam , Tanzania [44] . Of disease studies in humans or animals , none estimated disease incidence , and two studies [45] , [46] of livestock abortions met OIE definitions for either presumptive or confirmed cases . The remaining 5 animal abortion studies were serological investigations in individuals with history of abortions [34] , [47]–[50] . Of two surveys of cattle with abortions in northern Cameroon , one did not detect serological evidence of C . burnetii infection in any cattle [50] , while 3% in the other study were seropositive , compared to 7% among a non-random selection of cattle without abortions [34] . In South Africa , C . burnetii was found by smear microscopy in aborted calf fetuses at all of six cattle farms sampled [46] . In the Maradi Region of Niger , 32% of goats with previous abortions were seropositive , compared to 29% of non-randomly selected goats without a history of abortion [47] . Sheep with a history of abortion in Rabat , Morocco , were more likely than those with normal births to be seropositive for C . burnetii , 33% versus 15% ( p<0 . 01 ) [48] . In a survey conducted in five Tunisian governorates , 7% of sheep without past abortions were seropositive for C . burnetii compared to 12% of small ruminants with previous abortions [49] . Another Tunisian study found that 19% of small ruminants with a history of abortion had C . burnetii detected by PCR analysis of birth products or vaginal secretions [45] , and in South Africa , the pathogen was found by smear microscopy in aborted lamb fetuses from all of six sheep farms sampled [46] . Human cohorts comprising individuals with infective endocarditis in Sousse and Sfax , Tunisia , as well as Algiers , Algeria , have demonstrated C . burnetii as the causative pathogen in 1–3% of cases [51]–[53] . Two studies of febrile patients in Sousse , Tunisia , serologically identified acute Q fever in 2% and 9% of hospital admissions [54] , [55] . Q fever was responsible for 5% of patients with acute febrile illness hospitalized in Bobo-Dioulasso , Burkina Faso [56] as well as 3% of pediatric and 8% of adult admissions for severe febrile illness at two referral hospitals in the Kilimanjaro Region of northern Tanzania [11] . In two studies of patients admitted for community-acquired pneumonia in Yaoundé and Douala , Cameroon , 6% and 9% of persons aged >15 years had serologically-confirmed acute Q fever [57] , [58] . In these studies , Q fever was the third most common etiologic agent of pneumonia , after Streptococcus pneumoniae and Mycoplasma pneumoniae . At a major hospital in Cape Town , South Africa , C . burnetii was not found to be the cause of any pneumonia cases in a 92-patient cohort [59] . Among Nigerian cattle near Zaria , no difference in seropositivity was detected for cattle managed semi-intensively versus extensively [30] , [31] . In Cameroon , positive associations were found between seropositivity and cattle aged >2 years , female animals , those seen grazing with buffalo , and those for which the owner's ethnic group was recorded as Mbororo or ‘other’ when compared to Fulbe [33] , [60] . In the Egypt study linking human and animal populations , rural human residents were more likely to test seropositive than those in urban areas [27] . In the linked study from Chad , human Q fever serostatus did not correlate with the proportion of seropositive animals within respective nomadic camps , and camel breeders were at higher risk for Q fever seropositivity than cattle breeders [28] . In Ghana , children of illiterate mothers had a two-fold higher risk of seropositivity compared to those of literate mothers [41] . There was no association detected between C . burnetii seropositivity and HIV serostatus in pregnant Tanzanian women [44] . In the hospitalized patient cohort in northern Tanzania , there was no difference in prevalence of acute Q fever infection in HIV-infected compared to HIV-uninfected individuals [11] , and all cases of community-acquired pneumonia in the surveys of hospitalized patients in Cameroon were in HIV uninfected individuals [57] , [58] . No studies of risk factors for animal disease remained after quality assessment . No disease control studies were found by our search .
Superseded geographic or biological terminology may have caused us to inadvertently miss pertinent research . The already remote chances of communicating with authors of older manuscripts were complicated by the absence of electronic contact information . We excluded arthropod vector studies , but surveys of invertebrates and non-domestic animals may contribute to knowledge about C . burnetii transmission . Comparisons between studies and sub-regions were restricted by changes in diagnostic methods over time , frequently small sample sizes , and the low total number of studies . The low number of studies for each research question and the heterogeneity of these studies precluded a more extensive quantitative analysis of the epidemiology of C . burnetii in Africa . To our knowledge , this is the first systematic review of the epidemiology of C . burnetii in Africa from a ‘One Health’ perspective . Taken together , these findings suggest: 1 ) exposure to C . burnetii is a common finding in many animal host species across Africa , but seroprevalence varies widely by species and location , and the risk factors underlying this variability are largely unknown; 2 ) C . burnetii has been implicated as a cause of livestock abortion and could be responsible for substantial economic burdens , but more rigorous studies are required to determine this and other sequelae of disease in animals; 3 ) risk factors for human exposure to Q fever are poorly understood , but a more detailed understanding of how human exposure in different communities is linked with animal infection patterns and animal husbandry practices is clearly needed; and 4 ) Q fever accounts for a notable proportion of undifferentiated human febrile illness and infective endocarditis but studies describing other acute or chronic disease manifestations are scarce . The picture is complex , but the existing literature suggests that C . burnetii is found across diverse settings in Africa and presents a real yet underappreciated threat to human and animal health throughout Africa . | Coxiella burnetii is a bacterium that can cause acute and chronic fever illness and pneumonia in humans . It is also a known cause of abortion in livestock species , and is principally transmitted to humans through contact with infected animal birth products . With growing awareness of the over-diagnosis and misclassification of malaria as the cause of fever illnesses in the tropics , including Africa , there is increased interest in the role of non-malarial causes of fever , such as C . burnetii . We performed a systematic review of the published literature on the epidemiology of C . burnetii in Africa to consolidate knowledge and identify knowledge gaps regarding the extent of this infection in humans and animals and the risk factors for infection transmission . Few studies on prevalence of infection in humans and animals used random sampling strategies , and among these only two studied linked human and animal populations . C . burnetii appears to be a common cause of severe fever illness in humans , but population-level incidence estimates are lacking . The differential risks for C . burnetii infection and potential control strategies within the various animal husbandry systems in Africa remain largely unexplored . We conclude that C . burnetii is an underappreciated threat to human and animal health throughout Africa . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
]
| [
"bacterial",
"diseases",
"infectious",
"diseases",
"veterinary",
"diseases",
"medicine",
"and",
"health",
"sciences",
"biology",
"and",
"life",
"sciences",
"veterinary",
"bacteriology",
"q",
"fever",
"veterinary",
"science"
]
| 2014 | Epidemiology of Coxiella burnetii Infection in Africa: A OneHealth Systematic Review |
The potent vasoconstrictor peptides , endothelin 1 ( ET-1 ) and angiotensin II control adaptation of blood vessels to fluctuations of blood pressure . Previously we have shown that the circulating level of ET-1 is regulated through its proteolytic cleavage by secreted serine carboxypeptidase , cathepsin A ( CathA ) . However , genetically-modified mouse expressing catalytically inactive CathA S190A mutant retained about 10–15% of the carboxypeptidase activity against ET-1 in its tissues suggesting a presence of parallel/redundant catabolic pathway ( s ) . In the current work we provide direct evidence that the enzyme , which complements CathA action towards ET-1 is a retinoid-inducible lysosomal serine carboxypeptidase 1 ( Scpep1 ) , a CathA homolog with previously unknown biological function . We generated a mouse strain devoid of both CathA and Scpep1 activities ( DD mice ) and found that in response to high-salt diet and systemic injections of ET-1 these animals showed significantly increased blood pressure as compared to wild type mice or those with single deficiencies of CathA or Scpep1 . We also found that the reactivity of mesenteric arteries from DD mice towards ET-1 was significantly higher than that for all other groups of mice . The DD mice had a reduced degradation rate of ET-1 in the blood whereas their cultured arterial vascular smooth muscle cells showed increased ET-1-dependent phosphorylation of myosin light chain 2 . Together , our results define the biological role of mammalian serine carboxypeptidase Scpep1 and suggest that Scpep1 and CathA together participate in the control of ET-1 regulation of vascular tone and hemodynamics .
Vascular resistance of the mammalian circulation system is tightly regulated by many endogenous agents that influence the blood volume , and diverse functions of endothelium , vascular smooth muscle and myocardium . When the balance of these agents is disturbed , persistent systemic hypertension develops . Short regulatory peptides , endothelin-1 ( ET-1 ) and angiotensin II ( AII ) are recognized among the most potent vasoactive regulators . Through their interaction with cell surface receptors both peptides can modulate blood pressure by contracting arteries , or by induction or suppression of vascular wall remodelling . ET-1 also has mitogenic effects on vascular endothelium and smooth muscle [1] , stimulates the secretion of atrial natriuretic peptide ANP and aldosterone and inhibits the release of renin to counteract its effects [2] . The elevated ET-1 values have been previously observed in human vascular and cardiovascular disorders such as acute myocardial infarction , congestive heart failure , ischemia , atherosclerosis , hypercholestemia , systemic and pulmonary hypertension [3] . ET-1 deficient mice showed abnormal fetal development and haemodynamics [4] , whereas the overexpression of human ET-1 in mice caused vascular remodelling and endothelial dysfunction [5] , [6] . AII is another potent blood pressure-inducing and mitogenic peptide that belongs to the renin-angiotensin system . It is derived from the precursor , angiotensin I ( AI ) by angiotensin converting enzymes ( ACE or ACE2 ) . Inhibitors of AII receptors , as well as ACE inhibitors normalize the high blood pressure and decrease inward remodelling of arteries [7] . The bioavailability and potency of AII and ET-1 can be regulated through many factors such as alteration of receptor density and affinity , up- and down-regulation of peptide synthesis or release , enzymatic activation ( ACE and ACE2 for AII , ECE and MMP-2 for ET-1 [8] ) , or degradation ( neutral endopeptidase NEP for ET-1 [9]–[11] ) . Previously we have shown that circulating ET-1 is inactivated by lysosomal carboxypeptidase , cathepsin A ( CathA ) widely distributed in mammalian tissues ( reviewed in [12] ) . The majority of CathA in the cell is found in the lysosome but significant pool of the enzyme is also present at the cell surface and secreted outside the cell [12] . In vitro CathA rapidly inactivates ET-1 by converting it into biologically inactive des-Trp21-endothelin-1 [13] , [14] . CathA also hydrolyze the last residue of AI transferring it into angiotensin 1–9 ( A1–9 ) , which can be further converted into AII by ACE , but at much slower rate [15]–[17] . We reported that a gene-targeted mouse expressing enzymatically inactive CathA with a Ser190Ala mutation in the active site nucleophile [18] showed reduced degradation rate of ET-1 and significantly increased arterial blood pressure . At the same time , tissues of CathAS190A mice retained about 10–15% of the carboxypeptidase activity measured against ET-1 suggesting presence of parallel/redundant catabolic pathway ( s ) [18] . In the current work we tested whether the source of the complementary ET-1-degrading activity is a lysosomal serine carboxypeptidase 1 ( Scpep1 ) , a CathA homolog with previously unknown physiological function . Our results show that mice with a double CathA/Scpep1 deficiency ( DD mice ) demonstrate hypertension , increased ET-1-induced vasoconstriction and prolonged half-life of circulating ET-1 as compared to both wild type ( WT ) animals and those with single CathA or Scpep1 deficiencies , strongly supporting this hypothesis .
Mice with a combined deficiency of CathA and Scpep1 were obtained by intercrossing previously described CathAS190A and Scpep1−/− mouse lines , both in C57BL/6NCrl genetic backgrounds . Double heterozygous mice were crossed to obtain double homozygous CathAS190A/Scpep1−/− progeny , and their genotypes were confirmed by PCR of tail DNA ( Fig . S1 ) . CathAS190A/Scpep1−/− mice ( double-deficient mice , DD ) were viable and born in the frequency expected from Mendelian inheritance ( 16 of 307 ) indicating that combined deficiency of both enzymes does not cause embryonic lethality . CathAS190A/Scpep1−/− mice showed normal growth , were behaviourally indistinguishable from WT animals and could be bred to produce knockout litters . The amount Scpep1 mRNA measured by RT-q-PCR in aorta , hear and kidney tissues of Scpep1−/− and CathAS190A/Scpep1−/− mice ( Fig . S2 ) was below detection limit . Carboxypeptidase activity against ET-1 assayed in cultured AVSMC of CathAS190A mice was reduced to ∼10% of activity in WT mice whereas the activity in tissues of Scpep1−/− mice was reduced to ∼70% ( Fig . 1A ) . In the CathAS190A/Scpep1−/− mice carboxypeptidase activity was ∼6% of WT and significantly lower than that in CathAS190A mice , indicating that Scpep1 partially contributes to ET-1 hydrolysis ( Fig . 1A ) . The activity of Scpep1 against ET-1 was further confirmed when the AVSMC of CathAS190A/Scpep1−/− mice were transiently transfected with Scpep1-expressing plasmid [19] . The level of carboxypeptidase activity measured with ET-1 in the transfected cells was significantly higher then in non-transfected cells or cells transfected with control plasmid , coding for green fluorescent protein ( Fig . 1A ) and similar to that in the cells of CathAS190A mice , despite the modest transfection level of ∼5% that could be achieved in the primary AVSMC cultures . In contrast when we transfected AVSMC from WT mice with a lentiviral vector expressing shRNA for Scpep1 the carboxypeptidase activity against ET-1 in the cell homogenate in the transfected cells was reduced by ∼50% , consistent with that in the AVSMC from Scpep1−/− mice . In the cells transfected with CathA shRNA-expressing vector the activity was decreased by ∼90% and in the cells transfected with scrambled RNA constructs , not changed ( Fig . 1B ) . Finally , to test directly if Scpep1 has carboxypeptidase activity against ET-1 we have expressed the protein , carrying a His6 tag at the C-terminus ( Scpep1-His6 ) in HT1080 cells [19] . The secreted protein was purified until electrophoretic homogeneity by affinity chromatography on Ni-NTA resin followed by anion-exchange chromatography on Poros HQ resin ( Fig . S3 ) and its carboxypeptidase activity was assayed as above with 50 µM ET-1 as a substrate . We found that at pH 5 . 5 purified Scpep1-His6 was capable of cleaving the C-terminal Trp residue from ET-1 at a rate of 23 . 6 µmol/h per mg of protein ( Fig . 1C ) , i . e . close to that of purified CathA [20] . Lower activity was observed at higher pH of 6 . 5 and 7 . 5 ( Fig . 1C ) . The heart rate ( HR ) and blood pressure ( BP ) in WT , CathAS190A , Scpep1−/− and double-mutant CathAS190A/Scpep1−/− male mice was measured by radiotelemetry over a 3-day period . Then , the mice were challenged by a high salt diet ( 8% NaCl for 2 weeks ) with continuous measurement of BP and HR . The day ( Fig . 2A , C ) and night ( Fig . 2B , D ) levels of systolic BP ( SBP ) were significantly increased in CathAS190A animals as compared with WT both before and during a high-salt diet , whereas the BP levels in Scpep1−/− animals were similar to that of WT . Night SBP in CathAS190A/Scpep1−/− mice was significantly different from that of WT , Scpep1-deficient and CathA-deficient animals ( Fig . 2 and Fig . S4 ) . The HR values ( Fig . S5 ) and the parameters characterizing kidney function ( water intake , urine volume , urine sodium and urine creatinine levels , Fig . S6 ) were similar for all strains suggesting that the observed increase in SBP in DD mice relates to a vascular effect reflecting potential roles of Scpep1 and CathA in conversion of vasoconstrictive peptides . This hypothesis was further tested by measuring changes in BP in response to ET-1 and AI , the precursor of vasoconstrictive peptide AII . The changes in the diastolic and systolic BP ( ΔDBP and ΔSBP , respectively ) were calculated as the differences between the measured BP values and the basal values measured for the 30 min interval preceding the injection . To reduce the impact of the stress on BP caused by animal handling/injection , animals were receiving daily saline injections for 3 days prior to the experiment . The data ( Fig . 3 ) indicate that the BP response to the i . v . injections of ET-1 was significantly ( p<0 . 0001 ) dependent on animal genotype . The effect of ET-1 in Scpep-1-deficient mice was not significantly different from that in WT animals: there was no BP increase in response to the low ( 0 . 2 nmol/kg ) dose of ET-1 and similar increase in response to the high ( 5 nmol/kg ) dose . While CathA-deficient animals showed a higher response to ET-1 at the low dose as compared to WT or Scpep1-deficient animals only , DD animals showed a higher response as compared to all other groups of mice ( Fig . 3A , B ) . After injection of ET-1 at the high dose the BP in DD mice remained significantly elevated for at least 60 min , whereas in WT or single knockout mice it decreased already after 40 min ( Fig . 3C , D ) . The effect of AI on BP was similar in all animal groups ( Fig . S7 ) . The vasoreactivity of mesenteric arteries from the four groups of male mice was directly measured in ex-vivo tests . Isolated arteries were exposed to ET-1 and the precursor of AII , AI as well as to known vasodilators ( acetylcholine , ACh and sodium nitroprusside , SNP ) and vasoconstrictors ( norepinephrine , NE . We observed no differences in vessel reactivity in response to ACh , SNP or NE ( Fig . 4A , B , C ) as well as to AI ( Fig . S8 ) between the four groups of mice . Reactivity to ET-1 was higher for CathAS190A and Scpep1−/− than for WT mice ( Fig . 4D ) . The reactivity of vessels from CathAS190A/Scpep1−/− mice to ET-1 was significantly higher than that for all other groups of mice consistent with the in vivo data showing bigger increase of BP in CathAS190A/Scpep1−/− mice in response to ET-1 ( Fig . 4D ) . To verify at the molecular level if AVSMC from CathAS190A/Scpep1−/− mice have increased reactivity to ET-1 , we studied intracellular signalling events in these cells in response to ET-1 . ET-1 interacts with G-protein-associated endothelin type A ( ETR-A ) and type B ( ETR-B ) receptors on the surface of AVSMC . Activation of the receptors induces phospholipase C and increases the intracellular Ca2+ level leading to activation of myosin light chain kinase that phosphorylates myosin light chain ( MLC ) [21]–[24] . This causes contraction of myosin filaments and shrinkage of the cells . We therefore , compared the level of MLC phosphorylation in AVSMC before or after treatment with ET-1 for the 4 strains of mice . AVSMC cultured overnight in serum-free medium were treated with or without 100 nM ET-1 , harvested and analyzed by Western blot using antibodies against MLC-2 phosphorylated at Thr18 and Ser19 residues or against total MLC protein . MLC-2 phosphorylation was blocked by pre-treatment of the cells with the known pharmacological antagonist of ETR-A , BQ610 , and ETR-B antagonist , BQ788 , suggesting that this effect is dependent on ET-1 action on its receptors ( Fig . 5A ) . The cells from CathAS190A/Scpep1−/− mice had significantly higher level of pThr18/pSer19-MLC-2: ∼2 times higher than that in the control , CathA-deficient or Scpep1-deficient cells , and elevated basal levels of MLC-2 phosphorylation ( Fig . 5B ) . To determine the ET-1 degradation rate we injected mice in the tail vein with an ET-1 solution in saline at a dose of 0 . 1 nmol/kg BW . Fifteen minutes after injection mice were sacrificed and their lungs and aorta as well as blood were collected to measure the concentration of ET-1 by ELISA . Endogenous levels of ET-1 were measured in the animals injected with saline . Our data show that , 15 min after the ET-1 injection , its concentration in lungs ( Fig . 6A ) and aorta ( Fig . 6B ) of CathA-deficient mice was higher than that in the WT or Scpep1-deficient animals thus confirming our previous findings about the involvement of this enzyme in the ET-1 degradation . In tissues or plasma of DD mice the concentration of ET-1 was significantly higher than that in WT , CathA-deficient or Scpep1-deficient mice suggesting that in CathAS190A/Scpep1−/− mice the degradation rate of ET-1 is considerably reduced . No differences in endogenous circulating levels of ET-1 ( Fig . 6D ) were recorded .
Scpep1 was originally identified in rat aortic smooth muscle cells by screening for retinoid inducible genes [25] . Retinoids , natural and synthetic derivatives of vitamin A , block SMC proliferation and attenuate neointimal formation after vascular injury , presumably through retinoid receptor-mediated changes in gene expression . High transcript levels of Scpep1 were detected in kidney , lungs and heart . Scpep1 was localized to lysosomes by immunofluorescence , subcellular fractionation assays and mannose 6-phosphate receptor binding [19] , [26] . Scpep1 shows high similarity to other members of the serine carboxypeptidase family and , in particular , to CathA . Like CathA , Scpep1 has a cleavable signal peptide , N-linked glycans , the Ser-Asp-His catalytic triad and is proteolytically processed from a 55 kDa precursor into the 35 kDa and 18 kDa fragments [19] . Scpep1 gene-interrupted mice generated by us using a gene trap technology are fertile , have normal growth , normal clinical blood and urine parameters and did not have pathological changes in any tissue examined [19] . Later study by Lee at al . [27] reported that the Scpep1-null mice generated by replacing exons 1 and 2 of the Scpep1 gene with Neo cassette show a decrease in medial and intimal cell proliferation as well as in vessel remodelling following arterial injury . The same study also reported that a ∼50% knockdown of endogenous Scpep1 in mouse ASMC line showed dramatic decrease in serum-stimulated growth . This study did not identify a physiological substrate of Scpep1 , but the authors concluded that Scpep1 and CathA have distinct functions and “non-overlapping pools of substrates that function in cardiovascular homeostasis” . Our current data , however , provide evidence that both carboxypeptidases catabolize at least one common substrate , ET-1 . Mice devoid of both CathA and Scpep1 activities show significantly higher BP on both normal and high-salt diet or in response to systemic injections of ET-1 as compared to WT mice or those with single deficiencies of CathA or Scpep1 . ET-1 also causes higher constriction of mesenteric arteries from DD mice . Since the effects of other tested vasodilators and vasoconstrictors are similar , these results are consistent with increased sensitivity of arterial smooth muscle to ET-1 . Indeed , in cultured AVSMC from DD mice ET-1 caused significantly increased phosphorylation of MLC-2 as compared with the control , CathA-deficient or Scpep1-deficient cells . Finally , the degradation rate of ET-1 in the blood plasma or aorta and lung tissues was significantly reduced in DD as compared to WT , CathA-deficient or Scpep1-deficient mice . The cardiovascular effects of AI and AII concentration in mouse plasma ( S . Ernest , unpublished ) were similar in WT , CathA-deficient , Scpep1-deficient and DD mice . This contradicts previously proposed role of CathA in the generation of AII from AI [15]–[17] and suggests that in general ET-1 and AII are controlled by different sets of proteases . We cannot exclude , however , that CathA still may participate in AII regulation in specific tissues , such as heart atrium , where the rate of AI conversion to A1–9 by CathA constitutes ∼25% of that to AII by ACE [17] Our data indicate that in mouse tissues CathA is sufficient for inactivation of ET-1 , which justifies the apparent absence of phenotype in our line of Scpep1 KO mice . In contrast , Scpep1 activity is unable to fully compensate for the loss in CathA activity in the knock-in CathA-deficient mice that show elevated blood pressure [18] . In the absence of CathA , the Scpep1 activity becomes essential for degradation of ET-1 as demonstrated by induced BP and contractility of arteries in DD as compared to singe CathA KI mice . Importantly , CathA has also other functions non-overlapping with those of Scpep1 , such as activation of sialidase Neu1 in the lysosome [12] , regulation of elastogenesis through its function in elastin-binding protein complex [28] , [29] and inactivation of bradykinin [30] . Intravenous bolus injections of potent specific CathA inhibitors induced bradykinin-dependent diuresis [30] , however in our experiments we did not see a difference in the urine volume between WT and CathA-deficient mice . One possible explanation is that CathA KI animals could adapt to deficiency of CathA by reducing bradykinin production or the number of bradykinin receptors . The expression of Scpep1 in cardiovascular tissues can be effectively induced by retinoic acid , potentially providing a metabolic bypath to correct arterial hypertension attributed to a deficiency in ET-1 degradation in galactosialidosis patients with mutations in the CATHA gene [31] , [32] . Interestingly all-trans retinoic acid has been shown to inhibit pulmonary hypertension induced by monocrotaline in rats [33] , whereas human patients with idiopathic pulmonary arterial hypertension were shown to have reduced retinoic acid levels [34] . The anti-hypertensive effect of retinoic acid treatment was attributed to its ability to elicit growth-inhibitory signals in pulmonary artery smooth muscle cells and influence pulmonary vascular remodelling [34]–[36] , while our current data allow to propose that it may be also related to the induction of Scpep1 followed by increased degradation of ET-1 . Together , our results define a biological role of Scpep1 protein , and suggest that Scpept1 and CathA participate together in the control of ET-1 regulation of vascular tone and hemodynamics .
Generation of mice containing Ser190Ala point mutation in the CathA active site ( CathAS190A strain ) and those with the Scpep1 gene interrupted by gene-trap technology ( Scpep1−/− strain ) have been described [18] , [19] . In the Scpep1 gene-trap mouse β-galactosidase/neomycin phosphotransferase ( geo ) fusion gene was inserted into intron 7 of the Scpep1 gene resulting in deletion of downstream exons 8–13 encoding in particular the putative catalytic triad amino acids , Asp371 and His431 from the gene trap transcript . The amount of Scpep1 mRNA and protein measured by Northern and Western blots in liver , kidney , heart , brain spleen and lung tissues of Scpep1−/− mice [19] as well as the amount Scpep1 mRNA measured by RT-q-PCR in aorta , hear and kidney tissues ( Fig . S2 ) was reduced below detection threshold of the methods . Both strains were back-crossed for at least 5 generations to C57BL/6NCrl strain distributed by Charles River ( QC , Canada ) . Homozygous animals from each genotype were cross-bread to obtain the Scpep1-deficient , CathA-deficient , double-mutant and wild type mice . Mice were housed in an enriched environment with continuous access to food and water , under constant temperature and humidity , on a 12 h light∶dark cycle . Approval for the animal care and the use in the experiments was granted by the Animal Care and Use Committee of the Ste-Justine Hospital Research Center . 50 µl of PCR mixture contained 100 pmol of each primer , 0 . 2 mM dNTPs , 1 . 5 U taq polymerase ( Feldan , 9K-001-0002 ) and 100 ng of genomic DNA from clipped tail tips in 20 mM Tris ( pH 7 . 4 ) , 50 mM KCl , and 1 . 5 mM MgCl2 . Multiplex primers for detection of Scpep1 alleles were 5′-ATCCTCACACATGCAAAGCA ( Scpep1-F ) , 5′-TATTGGGCTGGAGTGGAGAC ( Scpep1-R ) and 5′- CCTGGCCTCCAGACAAGTAG ( Scpep1-trap ) and for detection of CathA alleles , 5′-GGTGGCGGAGAACAATTATG ( CathA-F ) and 5′-AACAGAAGTGGCACCCTGAC ( CathA-R ) . For Scpep1 allele genotyping , samples were denatured at 94°C for 2 min , followed by 35 cycles at 94°C for 15 s , 52°C for 15 s and 72°C for 1 min , with a final extension reaction at 72°C for 30 s . For CathA allele genotyping , samples were denatured at 92°C for 5 min , followed by 30 cycles at 92°C for 30 s , 56°C for 30 s and 72°C for 30 s , with a final extension reaction at 72°C for 5 min . Then the amplification product was digested with NdeI ( Biolabs , R0111S ) at 37°C overnight . Total RNA was isolated from mouse tissues using the Trizol Reagent ( Invitrogen 15596-026 ) according to the manufacturer's protocol and reverse-transcribed using random primers and QuantiTect Reverse Transcription Kit ( QIAGEN 205311 ) . Quantification of mouse Scpep1 mRNA was performed using an SsoFast EvaGreen Supermix with Low ROX ( BIO-RAD 172-5210 ) and the following set of primers: 5′- AGCAAGGGACCATTAAGTGC-3′ and 5′-GCTGAGTGGCCTCCTTGTAG-3′ . PCR conditions were as follows: 30 sec at 95°C , followed by 40 cycles of 5 sec at 95°C , 20 sec at 60°C , and 20 sec at 72°C . RPL32 mRNA was used as a reference control; the data were expressed as signal ratios between the test gene mRNA and RPL32 mRNA . Male CathAS190A mice and appropriate littermate controls were implanted with TA11PA-C10 radiotelemetry sensors ( Data Sciences International ) in the left carotid artery for direct measurement of arterial pressure and heart rate as described [37] , [38] . The transmitter was placed subcutaneously along the left flank . For basal measurements of mean day and night BP data were recorded continuously ( sampling every hour for 20 sec ) within 16 consecutive days and averaged for 12 h light and dark intervals . To measure changes in BP after ET-1 and AI injections data were recorded every 3 min for 2 h and averaged for 10-min consecutive intervals . At least 7 mice were studied for each genotype with the exception of WT mice for which only 5 mice were tested due to sudden death of 2 animals . Vessel reactivity ex vivo was analyzed as described [39] , [40] . Briefly , male mice were sacrificed at five months and their mesenteric arteries were isolated and mounted onto glass capillaries in an artereograph chamber filled with cold oxygenated Krebs solution ( 118 . 6 mM NaCl , 4 . 7 mM KCl , 1 . 2 mM KH2PO4 , 1 . 2 mM MgSO4 , 25 . 1 mM NaHCO3 , 26 µM EDTA , 0 . 18% glucose , 2 . 5 mM CaCl2 ) . The arteries were constantly perfused intraluminally with Krebs solution at 30 mmHg . After 45 minutes of equilibration vascular reactivity was measured in response to Norepinephrine ( Sigma A-0937 , 10−9–10−5 M ) , Acetylcholine ( Sigma A-6625 , 10−9–10−4 M ) , Sodium Nitroprusside ( Calbiochem 56538 , 10−9–10−4 M ) , Endothelin-1 ( American Peptide Company 88-1-10 , 10−11–10−8 M ) and AI , ( American Peptide Company 12-1-10 , 10−8–10−4 M ) . Drugs were added extraluminally with a 30 min washout period in between each drug , during which the arteries were able to re-equilibrate to a baseline . To test vasodilatation arteries were pre-contracted with NE to 70% of their equilibration diameter . At least 3 concentration response curves were conducted for each vessel and at least 6 animals were studied for each genotype . Combined tissues from 5–6 mouse aortas were minced in a DMEM containing collagenase type I ( GIBCO , 17100-017 , 3 mg/ml ) , trypsin ( Sigma T-1426 , 0 . 5 mg/ml ) , and DNAse type I ( Sigma D-4263 , 20 µg/ml ) , incubated at 37°C for 2 h , and centrifuged for 5 min at 1000 g and 4°C . The cells were resuspended in 10 ml of DMEM containing 10%FBS , 1% Antibiotic-Antimycotic ( GIBCO 15240-062 ) , 0 . 5% Fungizone ( GIBCO 15290-018 ) and maintained in 5% CO2 incubator at 37°C . The medium was changed every three days . After 3 passages 100% of cells were positive to VSMC marker , smooth muscle α-actin as assayed by FACS with A 2547 antibody ( Sigma ) . HT1080 cells stably expressing Scpep1-His6 [19] were cultured in DMEM with 0 . 05% FCS . Medium was collected three times every 48 h and subjected to ammonium sulfate precipitation . After dialysis against PBS , the Scpep1-His6 was purified by metal affinity chromatography on Ni-NTA agarose ( Qiagen ) as recommended by manufacturer . The eluate was dialyzed against PBS and subjected to HPLC anion exchange chromatography ( BiocadVision , Applied Biosystems ) by applying a step-wise gradient up to 500 mM NaCl in PBS . Purity of Scpep1-His6 was monitored by silver staining and Western blotting . Carboxypeptidase activity in cultured AVSMC was measured against 50 µM ET-1 as previously described using the method measuring the liberation rate of the C-terminal amino acid of the peptide [18] . Subconfluent AVSMC were transiently transfected or not with Scpep1-RGS-His-Tag [19] and pEGFP-C1 ( Clontech , Palo Alto , CA ) plasmids , mouse CTSA shRNA ( TF501716B/C ) Scpep1 shRNA ( TF505007A/B ) or non-effective 29-mer scrambled shRNA ( TR30015 ) cassette in pRFP-C-RS vector ( Origene Technologies ) using Effectene transfection reagent ( Qiagen ) at a ratio of 25 µl of Effectene to 1 µg of DNA . Forty eight hours after transfection ( 72 h for shRNA constructs ) confluent cells were harvested , homogenized in water by sonication and 50 µl of cell homogenate was mixed with 100 µl of 0 . 1 mM ET-1 solution and 50 µl of 100 mM sodium acetate buffer , pH 5 . 4 , and incubated for 30–180 min at 37°C . After addition of trichloroacetic acid ( Sigma T0699 , 3% final concentration ) proteins were removed by 5 min centrifugation at 12 , 000 g . The 190 µl aliquot of supernatant was mixed with 3 ml of 50 mM sodium borate buffer , pH 9 . 5 , containing 0 . 15 mg/ml of phthalic aldehyde and 1 mM of beta-mercaptoethanol ( Sigma , M-6250 ) and incubated at room temperature for 30 min . The fluorescence was measured at 340 nm excitation and 495 emission wavelength and concentration of released amino acids determined using a calibration curve established with 1–100 µM leucine . Carboxypeptidase activity of recombinant Scpep1-His6 was measured by the same method using 0 . 4–0 . 8 µg of the purified enzyme . AVSMC cultured in 100 mm dishes to confluent layer were incubated overnight in a serum-free DMEM , and treated for 5 min with 100 nM ET-1 . To test the pharmacological inhibition of the ET-1 receptors the cells were pre-treated for 30 min with 2 µM BQ610 ( EMD 203715 ) or BQ788 ( EMD 5223838 ) before stimulation with ET-1 . The cells were washed with ice-cold PBS , and lysed in RIPA ( RadioImmunoPrecipitation Assay ) buffer containing 50 mM Tris HCl , pH 7 . 4 , 150 mM NaCl , 1% NP-40 , 0 . 25% sodium deoxycholate , 0 . 1% SDS , 2 mM EDTA , 1 mM PMSF , protease and phosphatase inhibitor cocktails ( Roche 04693132001 and 04906837001 ) . Cell lysates were analyzed by Western blot using anti-phospho-Thr18/Ser19 myosin light chain 2 antibody ( Cell Signalling 3674 , dilution 1∶1000 ) or anti-myosin light chain 2 antibodies ( Cell Signalling 3672 , dilution 1∶1000 ) . Detection was performed with anti-rabbit IgG antibodies-HRP conjugate ( Cell Signalling 7074S ) , and the enhanced chemiluminescence reagent ( Thermo 32106 ) . Three to four month old mice with 25–35 g body weight ( BW ) were anesthetised with urethane ( 1 . 5 g/kg BW ) and injected into the tail vein with a solution of ET-1 in saline at a dose of 10 nmol/kg BW . Fifteen minutes post-injection , blood was collected in EDTA-coated tubes through cardiac puncture and immediately centrifuged to separate plasma . Aortas and lungs were dissected and rapidly frozen in liquid nitrogen . For peptide extraction , tissues ( 200 mg ) were homogenized in 1 mol/L CH3COOH/20 mM HCl . Plasma was supplemented with concentrated CH3COOH until the final concentration of 1 mol/L . Samples were boiled for 10 minutes and centrifuged at 20 , 000 g for 10 minutes . Supernatant was applied to a Strata C18-E column ( Phenomenex , RK-Sepcol-1 ) , washed with 3 volumes of 0 . 1% TFA in water , and peptides were eluted with 60% acetonitrile/0 . 1% TFA , lyophilized , and reconstituted in 0 . 1% TFA in DMSO . Quantitative assay of ET-1 was performed with an ELISA kit ( Enzo Life Sciences ADI-900-020A ) as described by the manufacturer . Statistical analysis has been performed using two-tailed paired t-test ( Fig . 1 , and 4 ) , Welch's modification of two-tailed unpaired t-test ( Fig . 5 , S2 and S8 ) and two-way repeated measures ANOVA ( Fig . 2 , 3 , and 4 ) tests using Prism Graphpad software . P-value of 0 . 05 or less was considered significant . Bonferroni post-hoc test was used to compare specific means , if significance was determined . The authors had full access to the data and take responsibility for its integrity . All authors have read and agreed to the manuscript as written . | Arterial blood pressure is regulated by small peptide hormones ( vasoactive peptides ) that cause contraction or relaxation of the arterial wall . The blood and tissue levels of these peptides are controlled by two mechanisms: through their synthesis and through their inactivation by the enzymes that are capable of cleaving them . Our results demonstrate that vasoactive peptide endothelin-1 , is inactivated by two homologous enzymes , lysosomal serine carboxypeptidase , cathepsin A and lysosomal serine carboxypeptidase 1 . We have developed a mutant strain of mice that do not produce both enzymes and found that these mice rapidly develop high blood pressure and show a reduced degradation rate of endothelin-1 . We also found that endothelin-1 causes higher contraction of arteries from mutant than from normal mice or mice that are deficient only in one of the two enzymes . Our mouse model provides insight into the functional engagement of lysosomal serine carboxypeptidases in pathophysiology of hypertension and may become a tool to explore whether induction of these enzymes would have any therapeutic value . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
]
| [
"medicine",
"biochemistry",
"hemodynamics",
"protein",
"metabolism",
"proteins",
"plasma",
"proteins",
"biocatalysis",
"vascular",
"biology",
"regulatory",
"proteins",
"biology",
"hypertension",
"metabolism",
"cardiovascular"
]
| 2014 | Serine Carboxypeptidase SCPEP1 and Cathepsin A Play Complementary Roles in Regulation of Vasoconstriction via Inactivation of Endothelin-1 |
Dengue virus infection is the most common arthropod-borne disease of humans and its geographical range and infection rates are increasing . Health policy decisions require information about the disease burden , but surveillance systems usually underreport the total number of cases . These may be estimated by multiplying reported cases by an expansion factor ( EF ) . As a key step to estimate the economic and disease burden of dengue in Southeast Asia ( SEA ) , we projected dengue cases from 2001 through 2010 using EFs . We conducted a systematic literature review ( 1995–2011 ) and identified 11 published articles reporting original , empirically derived EFs or the necessary data , and 11 additional relevant studies . To estimate EFs for total cases in countries where no empirical studies were available , we extrapolated data based on the statistically significant inverse relationship between an index of a country's health system quality and its observed reporting rate . We compiled an average 386 , 000 dengue episodes reported annually to surveillance systems in the region , and projected about 2 . 92 million dengue episodes . We conducted a probabilistic sensitivity analysis , simultaneously varying the most important parameters in 20 , 000 Monte Carlo simulations , and derived 95% certainty level of 2 . 73–3 . 38 million dengue episodes . We estimated an overall EF in SEA of 7 . 6 ( 95% certainty level: 7 . 0–8 . 8 ) dengue cases for every case reported , with an EF range of 3 . 8 for Malaysia to 19 . 0 in East Timor . Studies that make no adjustment for underreporting would seriously understate the burden and cost of dengue in SEA and elsewhere . As the sites of the empirical studies we identified were not randomly chosen , the exact extent of underreporting remains uncertain . Nevertheless , the results reported here , based on a systematic analysis of the available literature , show general consistency and provide a reasonable empirical basis to adjust for underreporting .
Dengue has become a major public health problem in many tropical and subtropical regions , with an estimate of 100–200 million dengue infections occurring each year in more than 100 countries , resulting in approximately 20 , 000 deaths [1] . It represents a significant economic burden to communities and health services in endemic countries , with a thirtyfold incidence increase in the last fifty years [2] . A variety of factors have created the ideal conditions for the expansion and distribution of dengue mosquito vector and viruses , including high rates of population growth , inadequate water , sewer , and waste management systems , rise in global commerce and tourism , global warming , changes in public health policy , and the development of hyperendimicity in urban areas , although it is difficult to estimate the contributions of each factor separately [3]–[9] . Southeast Asia has the world's largest incidence of dengue , with cycles of epidemics of increasing magnitude occurring every three to five years [6] , [10] . The WHO regions of Southeast Asia and Western Pacific represent most of the current global burden of dengue [11] , and account for most deaths [12] . All four dengue serotypes have been found in most countries of SEA [13] which means that one person can get up to four dengue infections . The risk of developing a more severe manifestation of dengue illness ( e . g . , dengue hemorrhagic fever ( DHF ) and dengue shock syndrome , DSS ) increases with subsequent infections [14] , [15] , which may explain the higher frequency of DHF and DSS in hyperendemic countries , such as those in SEA . Having accurate information about the human and economic burden of dengue is essential to inform policy makers and international donors , set health policy priorities , and make informed decisions about disease-control technologies and resource allocation . Estimating the total cases of symptomatic dengue infection is a critical step in calculating its economic and disease burden , but dengue incidence data are heterogeneous and incomplete . Most dengue episodes are identified using clinical diagnosis and existing epidemiological information , and , less frequently , clinical laboratory tests [13] . Passive surveillance systems require that identified episodes of dengue fever be reported , usually to the Ministry of Health ( MoH ) , but reported episodes do not consistently indicate severity of dengue . While passive surveillance systems , which rely on clinicians' reports incidental to their providing treatment , are appropriate to help detect dengue outbreaks promptly and examine long-term trends , they are not designed to estimate the real disease burden and usually underreport the total number of symptomatic dengue cases [13] , [16]–[20] . Limitations of passive surveillance systems include variations in both the system design and in the characteristics of surveillance implementation . Variations in system design include dengue case definition , clinical or lab-confirmed diagnosis , inpatient and outpatient reporting , reporting from sentinel or all hospitals and health services , public and private sector reporting , specific ages or dengue severity reported , and surveillance budget . The characteristics of surveillance implementation include reliance on health care professionals and laboratory staff , use of electronic or paper forms , and people's health-seeking behavior . Other sources of variation of dengue reporting rates include whether data is collected on epidemic or non-epidemic years , unrecognized or mild symptoms , the overall quality of the health care system , and the specific area of the country where incidence of dengue is measured ( e . g . , rural or urban ) [6] , [13] , [17] , [21]–[26] . While Singapore has recently implemented more sophisticated techniques to report cases and promptly respond to dengue epidemics [24] , [26] , and Malaysia has some enhanced capacity [24] , most countries in SEA have only passive surveillance systems [25] , making underreporting of dengue episodes a significant challenge in estimating disease burden . Based on a systematic literature review and using available data from surveillance systems , we estimated the average annual episodes of dengue by type of treatment in 12 countries from SEA ( 2001–2010 ) . We defined SEA for this study as consisting of the following 12 countries: Bhutan , Brunei , Cambodia , East-Timor , Indonesia , Laos , Malaysia , Myanmar , Philippines , Singapore , Thailand , and Viet Nam . We wanted to focus on a contiguous area with hyperendemic dengue and a reasonable amount of available data . We included all countries in the Association of Southeast Asian Nations , as in a previous cost-effectiveness study [27] , and added the bordering countries Bhutan and East-Timor due to their geographic proximity [28] . We obtained the incidence of symptomatic dengue by adjusting the reported episodes using an “expansion factor” ( EF ) , which corrected for underreporting . While total dengue episodes remain an area of considerable uncertainty , we used the best empirical data available to provide as accurate estimates as possible .
We combined various data sources to obtain our best estimate of the average annual episodes of dengue in 2001–2010 . The data included a systematic literature review of articles ( 1995–2012 ) that reported empirically derived EFs or the necessary data to estimate them , and available surveillance data on reported dengue episodes . An EF is the number by which reported cases need to be multiplied to obtain the most accurate estimate of the true number of episodes The goal was to obtain total episodes by identifying the following parameters: reported episodes , EFs for total ( EFT: total episodes/reported episodes ) , hospitalized ( EFH: total hospitalized episodes/reported hospitalized episodes ) , and ambulatory cases ( EFA: total ambulatory episodes/reported ambulatory episodes ) where possible , and last , the outpatient to inpatient ratio ( OP: IP ) to allow extrapolation from hospitalized episodes to total episodes , where necessary . Because the availability of country-specific data and the quality of published studies varied substantially , we used the best data available and extrapolated parameters based on a measure of health system quality and on assumptions about reporting of hospitalized episodes and OP∶IP ( discussed below ) . We obtained the reported cases of dengue from various sources , including surveillance data from country-specific MoH or department of statistics [29]–[34] and WHO regional offices [11] , [13] , . Cambodia was the only country considered in this study that only reported dengue episodes affecting patients of less than 15 years old [17] , [26] , [38] . Because dengue is an infectious disease , the number of cases varied considerably among years . To generate more stable estimates of the total projected cases of dengue , we considered the average reported cases in the last decade of available data ( 2001–2010 ) . An EF can be calculated as the analyst's best estimate of the total number of dengue episodes in a specified population divided by the episodes reported ( whether or not they actually were laboratory-confirmed dengue ) . A recent study of the economic impact of dengue in the Americas identified five studies that permitted estimating EFs for the reported episodes of dengue [16] . The EFs ranged from 1 . 6 inpatient dengue episodes reported for each hospitalized episode in Brazil ( 1996–2002 ) [39] to 28 episodes of dengue for each clinically diagnosed episode in Nicaragua ( 2005–2006 ) [40] . While there are considerable regional differences in the epidemiology of dengue between SEA and the Americas , in both regions dengue consists of the same four virus serotypes , is officially notifiable , and is considerably underreported [41] . The attack rates of DHF and DSS in SEA are approximately 18 times that of the Americas , with infants and children most affected [41] . Some authors have used EFs obtained from studies in the Americas to estimate the burden of disease in Asian countries [42]–[44] . Given the differences in epidemiology and surveillance systems , we think it is more appropriate to rely on studies from the same region . To implement this approach , we conducted a systematic literature review of articles published in the Web of Science and MEDLINE databases using the keywords “dengue” and “surveillance”; “dengue” and “capture recapture”; or “dengue” and “sensitivity” . We included articles published between 1995 and 2012 in English , Spanish , French , or Portuguese , and obtained a total of 1 , 676 articles . We then reviewed the titles and abstracts of these articles and found 48 that contained information relevant to the study of EFs for dengue in SEA . We examined these 48 studies plus 14 related articles that we had collected from previous literature reviews in full text , checking references for any additional articles that we could have missed in our search ( e . g . , national publications not included in international indexes ) . Eight new articles resulted from reviewing the references . We then filtered this literature , retaining studies that explicitly reported systematic data on EFs or included the necessary data to estimate them . The specific retention criteria were: ( 1 ) use of original , empirical data , ( 2 ) implementation of a scientifically valid approach , and ( 3 ) the external validity of the data gathered ( plausible patterns among age groups , geographic regions , years , and study sites ) . We complemented the literature review with surveillance data to estimate corresponding EFs [31] , [33] . We used original , empirically derived EFT and EFH , or the data needed to derive them , where available . Because studies that reported EF used various designs , sampling criteria , methods of analysis , study settings , time frames , among other aspects , for some countries we had to make assumptions about the rate of underreporting in hospitals and/or the share of ambulatory dengue episodes to derive an estimate of EF . We discuss these assumptions further in the next section . For countries with no original , empirical data , we relied instead on extrapolation of EFT based on quality of health care of each specific country . Because we only had a few empirical observations of EFs , we created a Health Quality Index ( HQI ) using principal component analysis , including standardized measures of five country-level variables ( Table 1 ) [45] , [46]: physicians density per 10 , 000 population ( 2005–2010 ) , mortality rate for children <5 years ( probability of dying by age 5 per 1 , 000 live births ) , neonatal mortality rate ( per 1 , 000 live births ) , percentage of births attended by skilled health personnel , and total health expenditures per capita ( US$ ) . We chose these variables because they were readily available for most countries , and we hypothesized that they represented an underlying factor of healthcare quality . A similar method to address underreporting , based on a measure of accessibility to health care , was used by Murray et al . [47] in recent estimates for burden of disease for 291 diseases and injuries . With dengue being a primarily urban disease , we think that this measure of quality provides a more appropriate measure specifically for dengue . We then extrapolated values of EFT to other countries by running a linear regression with the reporting rates ( RR = 1/EFT ) as the dependent variable and the HQI as an independent variable . Because we only had five observations for EFT , we also included two additional empirically-derived estimates from Colombia ( EFT = 9 . 0 ) [48] and Nicaragua ( EFT = 20 . 3 ) [40] in the regression , and added a dummy variable to address regional differences . We used RR instead of EF to reduce heteroskedasticity , as suggested by visual inspection of residuals versus fitted values and a substantially lower chi-square in the Breusch-Pagan test . We used RR to predict EFT and confidence intervals from the regressions , and converted the final estimates into EFs for clarity . We only considered empirically-derived EFT and excluded estimates based on expert opinion ( i . e . , Malaysia ) from the regression model . While most underreporting occurs in ambulatory settings , hospitalized episodes are also underreported as indicated by the empirical evidence in the region and elsewhere [16] . The main reason for underreporting in hospitals , even in well-funded health systems such as Singapore , appears to be under diagnosis , which may occur because of limited sensitivity of some diagnostic tests and cost constraints [49]–[51] , or because patients are not routinely tested . Other plausible causes include reliance on clinicians' reports [21] , particularly in private settings , or limited technical expertise [51] . To estimate EFH for countries where no empirically-derived data were available , we ran a regression with hospitalized dengue RR as the dependent variable and HQI as an independent variable . Because we found no significant correlation ( see results for further discussion ) , we made two assumptions: ( i ) the rate of underreporting of hospitalized cases was , on average , the same as the average underreporting for countries in the region with empirical data , and ( ii ) OP∶IP of dengue episodes was , on average , the same as the weighted average of all available empirical studies in the region . While our assumption that EFH is constant for countries with no empirical data ignores idiosyncratic characteristics of health systems , we think that it is a reasonable approach given the relatively low variability we found on underreporting of hospitalized dengue episodes reported in empirical studies in SEA and elsewhere [16] . Assumption ( ii ) was only necessary for the countries in which reported data came from outpatient and inpatient sources , i . e . , Brunei , Laos , and the Philippines . Last , we estimated the annual average of dengue episodes by type of treatment ( inpatient and outpatient ) and the EFA for each country , where applicable . Because the total cases of dengue remained an uncertainty , we conducted a probabilistic sensitivity analysis , simultaneously varying our parameter estimates based on available information . For countries with empirical data , ( 1 ) we estimated the range for EFT using a program evaluation and review technique ( PERT ) distribution with empirically derived EFT as the best estimate and either the range of empirically derived estimates as the lower and upper bounds where available ( Cambodia and Thailand ) , or alternatively , the 95% prediction interval derived from the regression analysis as the lower and upper bounds ( λ = 4 to approximate the shape of a Normal distribution ) . We used prediction intervals because we were predicting individual EFT for a specific country , and not the expected value of EFT for all subjects , and the standard errors differ in both cases . ( 2 ) For Cambodia and Thailand , we varied OP∶IP using a normal distribution based on the weighted average and standard deviation from country-specific studies in different years and/or sites [19] , [20] , [52] . For Vietnam , Indonesia , Singapore , and Malaysia , for which we did not have enough country-specific OP∶IP observations , we varied EFH using a PERT ( λ = 4 ) distribution with the country-specific empirical estimate ( or expert-based estimate for Malaysia ) as the best estimate , used 1 . 0 as the lower bound to be conservative ( i . e . all hospitalized episodes of dengue were reported ) , and the maximum EFH ( 3 . 4 ) from empirical studies among the 12 countries as the upper bound . For countries where no country-specific empirical data were available , we varied ( 1 ) EFT using a normal distribution with μ and σ based on predicted estimates from the regression analysis , ( 2 ) EFH using a PERT distribution ( λ = 4 ) with 1 as the lower bound , the average empirical estimate from all 12 countries as the best estimate , and the highest empirical EFH estimate for all countries as the upper bound , and last , ( 3 ) OP∶IP using a normal distribution with μ = weighted average and σ = weighted standard deviation based on all available empirical results from the 12 countries . As an additional sensitivity analysis , we used triangular distributions instead of the PERT distributions to see how the results varied . We computed 20 , 000 Monte Carlo simulations for each parameter using RiskAMP , version 3 . 20 [53] , which uses the Mersenne Twister random number generator . Iterations drew random values from the distribution of each input . We present results with 95% certainty level bounds .
Figure 1 shows the total dengue episodes and associated deaths reported in SEA from 1988 through 2010 . The three countries with the most reported episodes of dengue fever were Viet Nam , Thailand , and Indonesia with cumulative totals of 1 . 73 , 1 . 54 , and 1 . 43 million cases reported , respectively ( 1988–2010 ) . Together they represent about 75% of the total reported dengue episodes in the region . In contrast , Bhutan , East Timor , and Brunei , which reported dengue only since 2002 , summed 3 , 358 cases over the same time period . Figure 1 shows the combination of cycles of dengue epidemics in SEA , which peaked in 1998 and 2010 with 540 , 000 and 650 , 000 overall reported episodes , respectively , and an increasing trend of total reported episodes that reflects both a growing problem and better reporting . Total reported deaths peaked at 3 , 500 in 1998 , and as expected , were significantly correlated to the total number of dengue episodes ( r2 = 0 . 74 , p<0 . 001 ) . We found considerable variation in surveillance systems in SEA , for example , in the type of dengue case reported by severity , age groups , or type of treatment . Table 1 shows the demographic , health quality , and surveillance system characteristics of the selected countries in SEA [24] , [26] , [31] , [36] , [38] , [45] , [46] , [54]–[60] . We identified 11 published articles that reported original , empirically derived EFs or the data needed to derive EFs [17]–[20] , [52] , [61]–[66] , one study based on a systematic two-round Delphi process [67] , three empirical studies on dengue burden [43] , [68] , [69] , and seven studies that used EFs based on secondary analysis of published data or exclusively based on expert opinion [42] , [44] , [54] , [70]–[73] . Table 2 shows the main results from the literature review for EFs , or necessary data , in SEA . We extracted data from the articles using a template similar to Table 2 , with additional columns ( e . g . , date the article was reviewed , limitations ) . We did not consider secondary analysis of data . Although this study is an original research study and not a systematic review , we adapted relevant parts of the PRISMA check list and flowchart to our literature review ( Figure S1 , Table S1 ) [74] . As shown in Table 2 , we found high quality data to estimate EFs for six countries: Cambodia , Thailand , Viet Nam , Indonesia , Singapore , and Malaysia . Our estimates for Thailand were based solely on cohort studies , and we combined a cohort and a capture-recapture study for Cambodia . We combined cohort studies and clinical surveillance studies to obtain our estimates for Indonesia and Viet Nam . The EFT for Singapore was based on blood samples from a national health survey and EFH was derived by combining these data with reported data from a recent multisite longitudinal study . We used data based on expert opinion to estimate appropriate EFs for Malaysia [67] . While some important studies were not yet available when the expert workshop on dengue reporting in Malaysia took place ( e . g . , [17] , [19] ) , the country's EFT and EFH were estimated through a rigorous two-round Delphi process ( see Figure S2 for details ) [75] , [76] . Table 3 shows a summary of the parameters used , sources , assumptions , and specific calculations for each country , which we discuss below . Cambodia reports dengue episodes only among children <15 years old , but approximately 90% of the cases of dengue in Cambodia occur within this group [54] , and about 80% occur among children <9 years old [77] . Dengue case definitions are based on WHO guidelines , and do not require laboratory confirmation -only a sample undergo serological or virological testing [77] . Our EF estimates for Cambodia were based on a cohort study [19] and a capture-recapture study [17] both in Kampong Cham province . Because both were carefully designed studies , we obtained EFH and EFT using a weighted average based on total dengue episodes by cohort by year . We also combined Vong et al . 's [20] and Wichmann et al . 's [19] cohort studies and obtained a weighted average of OP∶IP of 6 . 0∶1 . Table 4 shows the estimated EFs and total cases by country . For the sensitivity analysis , we varied EFT using a PERT distribution based on the range of empirical estimates , as stated above , and OP∶IP using a normal distribution with μ = 6 . 0 and σ = 1 . 2 – the weighted average and standard deviation based on total dengue cases from both studies . Table 5 shows a summary of the distributions and parameters used in the sensitivity analysis for each country . Thailand uses the WHO case definition to report patients of all ages , and laboratory testing is commonly applied to all hospitalized cases . Our estimates for EFs in Thailand were mostly based on Wichmann et al . 's study [19] but we refined their estimates using data from a previous cohort study ( 1998–2002 ) [52] , [66] . Wichmann et al . compared dengue incidence in the cohort to reporting data from the national surveillance dataset , stratifying data by type of management ( inpatient and outpatient ) , year , and age group , and estimated an average EFH of 2 . 9 , and OP∶IP of 2 . 5∶1 . Using these estimates , the authors derived an EFT of 8 . 4 . Another robust 1998–2002 cohort study in Kamphaeng Phet [52] , [66] provided an estimate of OP∶IP . A weighted average based on dengue episodes by year between these studies gave us an OP∶IP of 2 . 7∶1 . Using detailed surveillance data on reported cases for years 2003–2009 which suggests that on average 79% of reported dengue corresponds to hospitalized cases [33] , we adjusted Wichmann's [19] EFs . We derived an EFA = 29 . 8 and EFT = 8 . 5 ( Table 4 ) . For the sensitivity analysis , we varied OP∶IP using the weighted average and standard deviation based on total dengue episodes reported by Anderson et al . [52] and Wichmann et al . [19] , and EFT using a PERT distribution based the range of empirically derived estimates [19] ( Table 5 ) . We obtained EFT for Viet Nam based on a children cohort study with active surveillance in Lon Xuyen ( 2004–2007 ) [64] . Tien et al . compared the average annual incidence rate of laboratory-confirmed dengue in the cohort with incidence data obtained from the national surveillance system for the same years , age groups , and region . Using the reported data , we estimated an average EFT = 5 . 8 . Until 2005 , only DHF and DSS were reported in Viet Nam; hence , we assumed that most reported episodes of dengue were hospitalized . Tien et al . reported an average rate of OP∶IP of suspected dengue of 0 . 8 . While the OP∶IP ratio in Viet Nam is probably lower than in other countries because hospitalization is required for all children with suspected dengue [64] , the very high proportion of cases that were hospitalized might have been an artifact of the study's procedures ( a prospective children cohort adjacent to the provincial hospital ) . Instead , we used the weighted average OP∶IP for all studies in the region ( 4 . 4 ) and obtained an EFH = 1 . 2 . Using active surveillance , Phuong et al . [62] found that there were 5 . 2 serologically confirmed dengue episodes for each patient diagnosed with dengue -although the accuracy of dengue diagnosis was less than 50% . This number provides an external validation of our EFT estimate for Viet Nam , since if all cases diagnosed were reported – which is not likely the case – we would expect about 5 . 2 laboratory-confirmed episodes of dengue for each reported episode . The EFs for Indonesia were based on two empirical studies [18] , [63] . Reporting DHF episodes within 24 hours following diagnosis is required by law in Indonesia , and Chairulfatah et al . [18] found 3 . 3 hospitalized episodes of DHF for each reported episode , with 50% of cases >14 years . Because the accuracy of diagnosis increases with severity [64] , we would expect DHF episodes to be reported more frequently than only acute dengue inpatient episodes , so we believe our estimate is rather conservative . Considering that Indonesia only reported inpatient dengue episodes , we combined this EFH with Porter et al . 's estimate of 2 . 3 episodes of dengue for every hospitalized episode [63] , and obtained an EFT of 7 . 6 . We estimated EFT for Singapore mainly based on an empirical study by Yew et al . [65] . Yew et al . found evidence suggesting that only one out of 23 dengue infections ( including symptomatic and asymptomatic ) were notified . Because Yew et al . did not provide information on the ratio of asymptomatic to symptomatic cases of dengue infection in Singapore , we obtained this ratio from a weighted average based on the total number of dengue infections by cohort-year from cohort studies in Indonesia , Thailand , and Viet Nam [63] , [64] , [66] . On average , 18% of dengue infections were symptomatic , so we derived an EFT of 4 . 1 . We estimated EFH using the OP∶IP ( 1 . 16∶1 ) ratio derived from data reported by Low et al . from the multicenter longitudinal Early Dengue Infection and Outcome Study ( EDEN ) in Singapore , 2005–2010 [68] , [69] . Last , we obtained from Carrasco et al . [43] that 56 . 5% of the total dengue episodes reported to the surveillance system were hospitalized patients . Carrasco et al . obtained this proportion using data from the Communicable Diseases Division of the MoH , the EDEN study , and the Adult Retrospective Dengue Study at Tan Tock Seng Hospital ( ARDENT ) . From these estimates , we derived an EFH of 3 . 4 , and an EFA of 5 . 0 . Singapore has strict legal requirements and incentives for reporting dengue , and a high quality surveillance system . Thus , an EFH of 3 . 4 may be an overestimate of underreporting . To be conservative , we used instead an EFH of 2 . 5 , equivalent to the average EFH from empirical studies as our best estimate and 3 . 4 as the upper bound in the sensitivity analysis . For the same reasons , we also used 1 . 0 as the lower bound for both EFH and EFT in the sensitivity analysis . We obtained the EFs for Malaysia from a recent study by Shepard et al . [67] that combined multiple data sources to refine the estimates of underreporting of dengue cases , including data from the MoH , private laboratories , previous literature , and a two-round Delphi process ( Figure S1 ) . The first round of the Delphi process took place during a workshop in Malaysia , where evidence was discussed among experts from public and private sectors , and academia . The second round was conducted some weeks later among the same group after analyzing results from the workshop , updating evidence , and adjusting the results for internal consistency . The results from the Delphi process suggested an EFT = 3 . 8 , EFH = 1 . 7 , and an EFA = 65 . 6 . The estimated EFs were conservative , since some important studies were not yet available for either round ( e . g . , [17] , [19] ) . Shepard et al . obtained a distribution of dengue episodes by type of treatment using 2009 data , which was used to update their estimates using the average reported cases in 2001–2010 . We think these EF estimates were as accurate as available evidence allowed at the time the Delphi panel took place . We varied EFT and EFH for Vietnam , Indonesia , Singapore , and Malaysia in the sensitivity analysis , as shown in Table 5 . We extrapolated EFT based on the country's quality of health care , defined by HQI , for countries where no country-specific empirical data were available . The five standardized country-level variables were internally consistent ( Chronbach's alpha: 0 . 92; Kaiser-Meyer-Olkin measure >0 . 68 for all variables ) and loaded to a single factor that accounted for most of the variability of the data ( Eigenvalue: 3 . 8 ) . Figure 2 shows the empirically derived reporting rates by country and the regression results with a 95% confidence interval ( R2 = 0 . 93 , HQI significant at p<0 . 01 ) for SEA . We predicted EFT for countries where no empirical data was available based on these regression results ( Table 4 ) . To check robustness , we also ran the regressions using estimates for countries in SEA only , and obtained similar results ( R2 = 0 . 91 , HQI significant at p = 0 . 01 ) . Regression results were also similar using EFT as the dependent variable in the regression specification ( all countries: R2 = 0 . 74 , HQI significant at p = 0 . 08; only countries in SEA: R2 = 0 . 83 , HQI significant at p = 0 . 03 ) . We found no significant correlation between HQI and EFH , which is possibly explained by the relatively low variability of underreporting of hospitalized dengue episodes and the few observations available . The average EFH for countries in SEA with empirical data was 2 . 5 , which was within the range of EFH estimates obtained from systematic empirical studies in Puerto Rico [78] , [79] ( all episodes 2 . 4; DHF: 2 . 9 ) and Brazil [39] ( 1 . 6 ) . We also obtained a weighted OP∶IP average of 4 . 4∶1 [19] , [20] , [52] , [63] , [64] , [68] . Based on assumption ( i ) EFH = 2 . 5 , we directly derived the average distribution of dengue episodes in 2001–2010 for Bhutan , East Timor , and Myanmar . Considering both assumptions , ( i ) EFH = 2 . 5 and ( ii ) OP∶IP = 4 . 4∶1 , we estimated EFA and the distribution of dengue episodes by treatment for the remaining three countries: Brunei , Laos , and Philippines . Table 4 shows a summary of the results: EFs by country and the average annual reported and estimated total dengue episodes by type of treatment ( 2001–2010 ) . Overall , there were on average 386 , 154 annual dengue episodes reported in SEA from 2001 through 2010 . Using our expansion factors , we projected that a total of 2 , 917 , 368 symptomatic dengue episodes ( 95% certainty level: 2 , 722 , 270–3 , 378 , 463; interquartile range: 2 , 915 , 658–3 , 149 , 257 ) occurring each year on average , of which 815 , 636 were hospitalized ( 95% certainty level: 715 , 326–983 , 735 ) and 2 , 101 , 732 ambulatory ( 95% certainty level: 1 , 871 , 480–2 , 534 , 739 ) episodes . We obtained an overall EFT in the region of 7 . 6 ( 95% certainty level: 7 . 0–8 . 8 ) dengue episodes for every reported episode . Last , as an additional sensitivity analysis , we did 20 , 000 Monte Carlo simulations using triangular distributions instead of PERT distributions , maintaining the same lower and upper bounds and best estimate for EFs . The results from were very similar . We obtained a 95% certainty level of 2 , 498 , 726–3 , 513 , 599 total dengue cases ( interquartile range: 3 , 012 , 551–3 , 265 , 965 ) , 676 , 098–1 , 023 , 528 hospitalized cases , and 1 , 930 , 568–2 , 668 , 726 ambulatory cases . The 95% certainty level of overall EFT was 7 . 2–9 . 1 .
Obtaining an accurate estimate of the total number of episodes is a critical step in the study of the disease and economic burden of dengue . Our analysis suggested that there is substantial underreporting of symptomatic dengue illness in SEA , with an average of only about 13 . 2% ( 95% certainty level: 11 . 4%–14 . 3% ) of all symptomatic dengue episodes reported to surveillance systems . Under-reporting is particularly a problem during inter-epidemic periods , while over-reporting ( or substantially less under-reporting ) might occur during epidemics . Undifferentiated fever due to dengue is indistinguishable for other viral fevers and even for DHF the differential diagnoses are very broad in the early febrile phase . But on balance , the overall effect is for all dengue to be underreported as evidenced by active surveillance studies of dengue in the region . We estimated a total of about 2 . 9 million annual dengue episodes occurring in 12 countries in SEA ( 2001–2010 ) , with an average ratio of OP∶IP of 2 . 6 ( 95% certainty level: 2 . 0–3 . 3 ) , which represent a serious burden to healthcare systems in the region . The strengths of our approach include our systematic procedures to identifying high quality empirical studies on EFs in SEA , our systematic inclusion of all the studies that met these standards , and our adjustment for the most salient site-level characteristics . In implementing our study , we conducted a systematic literature review and filtered the studies based on specific criteria . The resulting empirical EFs reflect the behavior of patients , health professionals , and the laboratory and public health systems in diagnosing , treating and documenting dengue . Because EFs are ratios of two measures ( projected and reported numbers of cases ) , they are more robust than raw numbers . High and low rates of dengue incidence tend to raise or lower both projected and reported numbers , without necessarily affecting the EF . The resulting empirical total EFs were relatively consistent , varying by a factor of only 3 . 1 from the lowest ( 4 . 1 ) to highest value ( 12 . 9 ) across all the countries . These similarities may reflect the health professionals shared experiences in training , professional conferences , publications , and guidance from the World Health Organization across SE Asia . Finally , while past studies have documented variations among dengue surveillance systems [6] , [24] , [26] , [38] , [80] , partly summarized on Table 1 , we were able to control for an important part of this variation . Because the completeness of dengue reporting was expected to reflect , in part , the quality of the health system overall , we controlled for this variation using a HQI . We expected higher quality health systems to have better reporting , and our regression results were consistent with these expectations . A relevant aspect of our estimates for total episodes of dengue is that they tend to smooth out geographic and time variation , since we used averages of reported cases in the last decade ( 2001–2010 ) and , in many cases , used weighted averages for parameters across studies and years . Even though our annual estimates do not reflect the actual idiosyncratic variability of total symptomatic dengue infections , they provide a more stable estimate of dengue burden in the region . While we believe our methods provide a reasonable empirical basis to estimate EFs and the total symptomatic episodes of dengue in SEA , several limitations must be acknowledged . First , the number of published empirical studies in SEA for estimating the total expansion factor is limited -only 11 . These studies varied in their methodologies ( e . g . , cohort studies , capture-recapture , hospital and health center surveillance , national surveys ) , age groups considered ( e . g . , cohorts of adult workers versus cohorts of children ) , types of study sites ( e . g . major public hospitals , rural health posts ) , or severity of dengue infection reported ( e . g . , dengue fever , DHF , DSS ) . Similarly , our regression relating RR to HQI was also based only on our best estimates for five countries in SEA and two in the Americas . When there is little information on a subject , each new piece can make a considerable difference . For example , the new available evidence estimating EFT for Malaysia , based on expert opinion , seems to be conservative [67] . As more empirical studies on underreporting become available , and surveillance systems improve their efficacy , EF estimates will be more accurate . Second , the specific study locations and age groups were generally ones to which the researchers had access . They were generally not randomly selected and are not necessarily representative of the country as a whole . For example , the specific studies in some countries may suggest a lower proportion of ambulatory cases relative to hospitalized cases , as might be the case of Indonesia , were the OP∶IP of 1 . 3 may be explained because textile workers in Bandung [63] probably have higher income and better healthcare access than the average person in that country . We also found other empirical studies where this ratio may seem too high [19] , [81] . Third , to the extent that site-to-site variation remains a factor , our ability to adjust for it was limited to our single variable , HQI . Also , the range we used for EFT in the sensitivity analysis considered only between-country and not within-country heterogeneity , which resulted in narrower ranges for our estimates of total episodes of dengue than if we had been able to include within-country heterogeneity . Fourth , we found no empirical or high quality studies for six of the 12 countries included in our estimates of dengue episodes . We addressed this constraint by extrapolating data from countries with empirical studies , but could not control for other factors , such as characteristics of national healthcare systems , the level of dengue awareness , or various other relevant factors , such as virus serotypes , rainfall , or global commerce and tourism [82] . As expected , the 95% certainty level for overall EFT in countries with empirical studies ( 6 . 8–8 . 1 ) was much narrower than for countries where we extrapolated EFT ( 6 . 4–13 . 6 ) . Fifth , some evidence suggests that the rate of underreporting varies by the severity of dengue symptoms , with reporting increasing for more severe dengue [39] . This evidence may imply that the most modest cases would have the highest degree of underreporting . Due to data limitations , we were not able to categorize numbers of reported cases by dengue severity . Despite all these limitations , we believe that adjustment for underreporting of dengue is critical to estimate the true economic burden , and we sought to make the best adjustments possible with available data . Estimating the rate of underreporting of dengue with accuracy is a very complex task , particularly distinguishing the factors that drive underreporting to surveillance systems . More accurate estimates of the rate of underreporting of dengue would require a better understanding of the epidemiology of dengue . We think that long-term nationally representative cohort studies that could factor in a wide range of variables related to healthcare systems ( e . g . , facility types , public and private sectors , number of physicians , specific lab tests used , dengue definition , diagnosis , healthcare access and coverage ) , geography ( e . g . , rural and urban population , altitude , latitude , rainfall ) , virus ( e . g . , dengue serotypes and genotypes ) , vectors ( e . g . , vector control activities , public awareness campaigns ) , and dengue sequelae ( e . g . , severity of disease , long-term symptoms , duration and intensity ) would be the ideal source of data . However such studies require considerably more time and resources than alternative designs , such as regional or local cohort studies , capture-recapture studies , or Delphi panels . By generating better estimates of EFs , this paper will contribute towards a better understanding of underreporting of dengue episodes and improving regional and country-specific estimates of the economic and disease burden of dengue in SEA [83] . While this study was focused on dengue in SEA , analogous principles apply to other regions of the world and other diseases reported through health information and surveillance systems . Estimating EFs is an important middle step towards estimating the economic burden of disease , and the cost-effectiveness of vaccines and other preventive and curative approaches . | Dengue is the most common disease transmitted by a mosquito , with about 100–200 million infections occurring each year in more than 100 tropical and subtropical countries . Policy-makers require accurate information about the number of symptomatic dengue episodes to make informed decisions concerning dengue control strategies . But dengue is usually underreported by national surveillance systems . Through a systematic literature review and analysis of empirical research , we estimated the rate of underreporting and the average annual dengue episodes by treatment ( hospitalized and ambulatory ) in 2001–2010 , for 12 countries in Southeast Asia . We found an average reporting rate of 13 . 2% of the total symptomatic dengue episodes in the region , leading to an expansion factor of 7 . 6 for converting reported cases into estimated actual cases . While we focused in Southeast Asia , analogous principles apply to other regions of the world , and other diseases reported through surveillance systems . Estimating the total episodes of dengue is a critical step in studying the economic and disease burden of dengue fever , and the cost-effectiveness evaluation of dengue control and prevention strategies . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
]
| [
"medicine",
"infectious",
"diseases",
"socioeconomic",
"aspects",
"of",
"health",
"public",
"health",
"and",
"epidemiology",
"epidemiological",
"methods",
"epidemiology",
"global",
"health",
"dengue",
"fever",
"neglected",
"tropical",
"diseases",
"spatial",
"epidemiology",
"economic",
"epidemiology",
"public",
"health"
]
| 2013 | Use of Expansion Factors to Estimate the Burden of Dengue in Southeast Asia: A Systematic Analysis |
mRNA half-lives are transcript-specific and vary over a range of more than 100-fold in eukaryotic cells . mRNA stabilities can be regulated by sequence-specific RNA-binding proteins ( RBPs ) , which bind to regulatory sequence elements and modulate the interaction of the mRNA with the cellular RNA degradation machinery . However , it is unclear if this kind of regulation is sufficient to explain the large range of mRNA stabilities . To address this question , we examined the transcriptome of 74 Schizosaccharomyces pombe strains carrying deletions in non-essential genes encoding predicted RBPs ( 86% of all such genes ) . We identified 25 strains that displayed changes in the levels of between 4 and 104 mRNAs . The putative targets of these RBPs formed biologically coherent groups , defining regulons involved in cell separation , ribosome biogenesis , meiotic progression , stress responses and mitochondrial function . Moreover , mRNAs in these groups were enriched in specific sequence motifs in their coding sequences and untranslated regions , suggesting that they are coregulated at the posttranscriptional level . We performed genome-wide RNA stability measurements for several RBP mutants , and confirmed that the altered mRNA levels were caused by changes in their stabilities . Although RBPs regulate the decay rates of multiple regulons , only 16% of all S . pombe mRNAs were affected in any of the 74 deletion strains . This suggests that other players or mechanisms are required to generate the observed range of RNA half-lives of a eukaryotic transcriptome .
The steady state levels of messenger RNAs ( mRNAs ) are determined by both their synthesis and decay rates [1] . Moreover , decay rates determine the time required for a new steady state to be reached after changes in transcription , and thus contribute to shaping dynamic changes of mRNA levels [2]–[5] . mRNA decay rates are transcript-specific [6] , [7] , and vary over a range of up to 100-fold [8] . In the budding yeast Saccharomyces cerevisiae , mRNA half-lives vary from a few minutes to over 3 hours ( with medians between 12 and 23 minutes depending on the method used for their determination ) [9]–[11] . The fission yeast Schizosaccharomyces pombe displays similar variations , with a median of 30 to 60 minutes [10] , [12] . mRNAs in mammalian cells are typically longer-lived , with half-lives ranging from less than 20 minutes to several days [13]–[16] . Eukaryotic mRNAs are protected from exonuclease degradation by the 5′ methylated guanosyl cap and the 3′ poly ( A ) tail , which is coated with poly ( A ) -binding protein . In the most common pathway , degradation starts with the removal of one or both of these protective structures , followed by digestion through the action of 5′→3′ or 3′→5 exonucleases . The enzymatic activities associated with cytoplasmic mRNA decay are performed by a small number of protein complexes , most of which are conserved from yeast to humans . Deadenylation is carried out by three different deadenylases ( Ccr4-Not , PAN2/PAN3 and PARN ) , decapping by the DCP1/DCP2 heterodimer , 5′→3′ degradation by the Xrn1 exonuclease , and 3′→5′ degradation by the cytoplasmic exosome [8] , [17] . Xrn1-mediated 5′→3′ decay appears to be the dominant pathway in S . cerevisiae [18] , [19] . The different steps of mRNA decay ( deadenylation , decapping , and exonuclease degradation ) take place with transcript-specific rates , and thus contribute to determine the overall decay rate [20] , [21] . The stability of a specific transcript is at least partly controlled by cis-acting sequences , which are frequently – but not always – located in 3′ untranslated regions ( UTRs ) [22] , [23] . These sequence motifs act by binding to sequence-specific RNA-binding proteins ( RBPs ) , which in turn modulate the interaction of the mRNA with elements of the core degradation machinery ( see [8] for a review ) . For example , many mammalian mRNAs contain sequences called AREs ( AU-rich Elements ) in their 3′ UTRs that make them unstable . Although the exosome is required for the rapid degradation of these mRNAs , it cannot recognize them on its own [24] . Instead , a number of ARE-binding proteins associate with these sequence elements and recruit components of the core degradation machinery , including decapping enzymes , the Ccr4-Not deadenylase complex , the exosome and Xrn1 [25] . Similarly , the budding yeast protein Puf5 , a member of the pumilio family of RNA-binding proteins , recruits the Ccr4-Not complex and components of the decapping complex to mRNAs [26] . Other ARE-binding proteins , such as HuR , stabilize their targets . Although the exact molecular mechanisms are unclear , this effect could be mediated through protection from miRNA-mediated degradation [27] . However , many unicellular organisms lack the machinery for the production of small RNAs ( such as S . cerevisiae ) [28] or have it but do not use it for the control of mRNA stability ( S . pombe ) [29] . It is becoming apparent that mRNA degradation and transcription are intimately linked . This was suggested by early work that showed that mutants in the S . cerevisiae dcp1 gene ( encoding a component of the decapping complex ) affected RNA stability without causing changes in RNA levels , possibly indicating compensatory changes in transcription [30] . More recently , several studies of RNA synthesis and decay rates have revealed widespread decreases in transcription rates in response to the inactivation of multiple RNA decay pathways such as the Ccr4-Not complex , Xrn1 , and the exosome [10] , [31] , [32] . Xrn1 appears to have a key function in this feedback by directly stimulating transcription [31] , [32] . The fission yeast Schizosaccharomyces pombe provides an attractive model for the study of posttranscriptional regulation in eukaryotic organisms . In addition to the major pathways described above , S . pombe contains decay-related components that are present in higher eukaryotes but not in S . cerevisiae , such as cytoplasmic poly ( A ) polymerases [33] , a poly ( A ) -specific ribonuclease ( PARN ) [34] and a poly ( U ) polymerase-dependent decay pathway [34] . It is still unclear how the specificity of decay rates is determined . Although RBP-mediated recruitment of the decay machinery can modulate the decay rates of individual mRNAs , it is not known if the large range in mRNA half-lives can be explained exclusively by this mechanism . To investigate this issue we performed gene expression profiling for 86% of S . pombe deletions in non-essential genes encoding RBPs . We found 25 strains that showed significant changes in RNA levels , affecting between 4 and 104 mRNAs . In addition , we identified 4 strains with defects in splicing . The potential targets of these RBPs had common properties , such as being coexpressed in response to stress , or encoding proteins with similar localization or involved in the same pathway . Unexpectedly , only 16% of all S . pombe mRNAs showed changes in levels in at least one of the 74 strains tested . This suggests that the action of these RBPs is not sufficient to explain the large range of mRNA half-lives in fission yeast , and that additional mechanisms may be required .
We focused on proteins that contained RNA-binding domains that might confer sequence-specificity , although we also investigated a few proteins with catalytic activities on RNA ( such as nucleases ) . Table 1 lists the 16 domains we selected with their PFAM references , and Table S1 presents all the selected proteins . The most common domains were the RRM ( RNA recognition motif , present in 72 proteins in fission yeast ) , PUF ( Pumilio Family RNA binding repeat , in 9 proteins ) , several zinc finger domains ( a total of 20 proteins ) , the KH ( K-Homology domain , 8 proteins ) and the G-patch domain ( 8 proteins ) . Seven proteins had domains from more than one family , and a total of 136 predicted proteins contained at least one of the selected domains . Of the genes encoding these proteins , 47 were essential , 86 were non-essential , and for the remaining 3 there was no information or conflicting data [35] , [36] . Therefore , the fraction of essential genes encoding RBPs in S . pombe is 34 . 5% , which is significantly higher than the overall percentage of essential genes ( 26% for all genes , p-value 6 . 5×10−3 ) . For the majority of the strains we used the Bioneer gene deletion library [35] ( see Table S11 for full details ) . The correct deletion of the genes was assessed from the microarray data and/or by gene-specific PCR . We confirmed the deletion was present in 74 genes out of our 86 target genes ( 86% ) , while in 12 strains the gene was not deleted ( Table S1 ) . Changes in stability are expected to affect mRNA steady state levels . Therefore , to identify RBPs with a potential role in decay , we used custom-designed oligonucleotide microarrays to compare the transcriptome of each of the 74 deletion strains to that of isogenic wild type cells grown under vegetative conditions . The microarrays contained two probes for every annotated S . pombe coding sequence as well as for 496 long non-coding RNAs and 1 , 491 introns . We initially carried out two independent biological replicates for each strain , and performed a third repeat for those that showed changes in the first two experiments . We used a robust statistical method to identify differentially-expressed genes ( Significance Analysis of Microarrays , or SAM ) [37] . A total of 25 deletions caused significant changes in gene expression , with the numbers of affected genes ranging from 4 to 104 ( Figure 1 ) . The majority of strains showed mostly up-regulated genes or both up- and down-regulation , while only few strains displayed predominantly down-regulation ( Figure 1 ) . Complete lists of affected genes and their corresponding enrichment analysis are presented in Tables S2 , S3 , and S4 . To validate our approach , we initially focused on those genes for which genome-wide transcriptome data was available . The zfs1 gene codes for a zinc-finger protein of the tristetraprolin family [38] and has been extensively studied at the genome-wide level . We found that zfs1 deletion caused up-regulation of 69 genes , which showed a highly significant overlap with the 60 genes identified in a published study [39] ( Figure S1A ) . In fission yeast vegetative cells a subset of meiotic mRNAs are bound to a protein called Mmi1 , which targets them for degradation by the nuclear exosome in a process that requires Pab2 ( a nuclear poly ( A ) -binding protein ) and Red1 ( a Zinc finger-containing protein ) [40]–[42] . pab2 mutants showed up-regulation of 33 coding genes , while red1 mutants overexpressed 63 mRNAs . The majority of mRNAs affected by the pab2 deletion were also overexpressed in red1Δ cells , and the effect was stronger in the latter ( average induction of 5 . 13 , compared to 2 . 8 in pab2Δ ) . This suggests that both proteins regulate the same group of mRNAs , although the function of Red1 appears to be more important . Both sets of mRNAs showed strong overlap with mRNAs physically associated with Mmi1 ( Caia Duncan and Juan Mata , unpublished data ) and with published microarray data on pab2 and red1 mutants [41] , [42] ( Figure S1B and S1C ) . Pab2 is also involved in the processing of snoRNAs [43] . The presence of long ncRNA probes in our microarray platform allowed us to investigate if Pab2 and Red1 are involved in the regulation of other non-coding transcripts . Indeed , we found 18 ncRNAs overexpressed in pab2Δ , and 35 in red1Δ ( Figure 2A and Table S5 ) . As was the case with coding genes , almost all ncRNAs up-regulated in pab2Δ were also induced in red1Δ ( Figure 2B and Table S5 ) . Some of these ncRNAs were overexpressed more strongly than most coding targets , suggesting that they may be functionally important . These results demonstrate that our strategy can identify both known targets and additional functions of RBPs . While this manuscript was in preparation , a transcriptome analysis of red1Δ mutants using high throughput sequencing was published [44] . This study reported overexpression of large numbers of ncRNAs in red1Δ cells , which showed a highly significant overlap with our data ( Figure S1D ) . Below we discuss a few notable examples of RBPs identified in the screen . scw1 encodes an RRM-containing protein and is essential for correct cell wall structure and cell separation . Mutants in scw1 display multiple septa , indicating a defect in cell separation , but the molecular basis for this phenotype is unknown [45] , [46] . scw1 deletion caused the overexpression of 76 genes , which were enriched in periodically-expressed genes ( Figure 2C and 2D ) as well as in genes encoding proteins required for cell wall organization and cytokinesis . For example , the induced genes included gas1 , gas2 and gas5 , encoding three 1 , 3-β-glucanosyltransferases [36] , bgl2 , which codes for a glucan exo-1 , 3-β-glucosidase [36] , pmk1 , encoding a MAP kinase that regulates cell wall integrity [47] , and rho2 , which codes for a GTPase that regulates cell wall alpha-glucan biosynthesis [48] . Our results suggest that overexpression and/or mis-expression of enzymes involved in cell wall biosynthesis are responsible for the cell separation phenotype of scw1 mutants . Mutations in scw1 supress defects in the SIN pathway , which regulates cytokinesis in S . pombe [45] , [46] . Moreover , Scw1 is a target of the Sid2 kinase , a key effector of the SIN pathway , suggesting that the SIN pathway may regulate Scw1 function [49] . The putative targets of Scw1 included mob2 , which encodes the regulatory subunit of Sid2 , suggesting that the interactions between the SIN pathway and Scw1 may be complex . In addition , scw1 mutants are released from a cdc25 G2 cell cycle block with very poor synchrony , possibly indicating mitotic defects ( JM and AH , unpublished data ) . It has been reported that scw1 mutants fail to arrest in mitosis in nda3 cold-sensitive mutants ( encoding β-tubulin ) , probably because mutations in scw1 lead to microtubule stabilization [45] . All together , these results suggest that Scw1 may have additional roles in cell cycle control . rpm1 ( also called par1 ) encodes a protein with a predicted exonuclease II domain that localizes to mitochondria [50] , [51] . Rpm1 is essential for growth in non-fermentable carbon sources , and rpm1Δ cells are defective in the processing of mitochondrially-encoded transcripts that encode key components of the respiratory chain , resulting in the accumulation of their RNA precursors [50] . rpm1 deletion caused reduced expression of a set of 38 nuclear-encoded genes , most of which encoded proteins localized to the mitochondrial envelope , including multiple components of the F0 and F1 ATPases and the cytochrome c oxidase complex ( Figure 2E ) . These genes showed a significant overlap with those under-expressed in cells lacking the m-AAA protease , which is involved in the processing of mitochondrial proteins [52] . Although the effect of Rpm1 on these mRNAs is likely to be indirect ( given that , as an exonuclease , its deletion would be expected to lead to overexpression of its targets ) , it appears to be posttranscriptional ( see below ) . The fact that two different mutations that affect the production of key mitochondrial protein complexes cause down-regulation of nuclear-encoded RNAs suggests the existence of a checkpoint mechanism that monitors the formation of respiratory complexes and responds by destabilising mRNAs encoding components of these complexes . SPAC17H9 . 04c codes for a protein with two zinc finger domains and an RRM , which is localized to the nucleolus and cytoplasmic dots [51] . SPAC17H9 . 04cΔ cells overexpressed 89 genes , ∼44% of which encoded proteins localized to the nucleolus and/or involved in ribosome biogenesis ( Figure 2F ) . These included 10 U3 snoRNA-associated proteins , 4 nucleolar ATP-dependent helicases , and several rRNA-modifying enzymes . In addition , mRNAs encoding proteins involved in rRNA processing , but not localized to the nucleolus , were also induced , suggesting that these genes define a novel regulon involved in rRNA maturation . mug187 encodes a predicted protein containing two RRMs that is induced during meiosis [53] and in response to multiple stress conditions , including cadmium treatment , heat shock and osmotic shock [54] . A total of 66 genes were induced in mug187 mutants ( grown in the absence of stress ) , which were enriched in genes specifically induced in response to cadmium , and in genes encoding enzymes required for serine and sulphur aminoacid biosynthesis . Several mutants showed up-regulation of ncRNAs ( Table S5 ) . Of particular interest was pan3Δ , which encodes a subunit of the Pan2/Pan3 complex . This complex has poly ( A ) -specific ribonuclease activity and is thought to regulate poly ( A ) tail length . Our results indicate that Pan2/Pan3 is required for the down-regulation of stress genes ( Table S3 ) and ncRNAs ( Table S5 ) . This is , to our knowledge , the first indication that this complex is involved in the degradation of ncRNAs . As discussed above , mRNA half-lives are transcript-specific and vary over a range of more than a hundred fold . To evaluate the contribution of non-essential RBPs to this phenomenon we quantified the fraction of genes whose expression was affected in at least one RBP mutant . A total of 816 genes were differentially expressed in at least one of the 74 strains we characterised ( 529 only up-regulated , 287 only down-regulated , and 46 induced in some strains and repressed in others ) . These genes represent only 16 . 2% of all coding sequences . This result suggests that even though non-essential RBPs influence the stability of several coherent sets of genes , their function is not sufficient to explain the large range in decay rates of the fission yeast transcriptome . Our array platform contains probes for 1 , 491 introns ( out of 4 , 730 in the S . pombe genome ) , allowing the identification of strains with deficient splicing . Splicing defects are expected to lead to the accumulation of pre-mRNAs and thus to overexpression of intronic regions . Only a few hundred introns were detected in most experiments , consistent with their short half-lives and general low levels ( see below ) . We observed splicing defects in strains that carry mutations in genes encoding predicted components of the core splicing machinery . usp102 encodes a U1 snRNP-associated protein that contains an RRM , and SPBC18H10 . 07 , a U1-type zinc finger predicted to bind to the U1 snRNA [36] . usp102 deletion had a modest effect with 25 overexpressed introns , while SPBC18H10 . 07 inactivation led to the accumulation of 268 introns ( Table S6 ) . In both cases , exonic probes of the corresponding genes did not detect any overexpression , suggesting that the increase in introns was due to the accumulation of unspliced pre-mRNAs , and not to a general increase in transcript abundance ( Figure 3A ) . Both sets of introns overlapped extensively , suggesting that both proteins regulate the splicing of the same mRNA set . We also identified novel proteins with a function in splicing . SPAC30D11 . 14c , which encodes an uncharacterised RBP that contains a KH domain [36] , overexpressed 50 introns ( Figure 3B and Table S6 ) . We performed GO-enrichment analysis of the corresponding genes , but could not find any specific enrichment . However , those introns overlapped significantly with those identified in usp102 and SPBC18H10 . 07 mutants ( Figure 3B ) , suggesting that they may represent a set of introns whose splicing is sensitive to decreased efficiency of the splicing machinery . The increase in RNA levels detected by intronic probes could also be the result of a failure to degrade spliced introns ( rather than an accumulation of unspliced pre-mRNAs ) . To address this possibility we performed RNA-seq experiments with mutants in SPBC18H10 . 07 and SPAC30D11 . 14c . We expected that a splicing defect would lead to an increase in both reads mapping to introns and to exon-intron junctions , whereas a defect in intron turnover would cause an accumulation in intronic reads but not in exon-intron junctions ( Figure S2A ) . Both mutants showed a build-up of intronic reads in specific sets of introns that overlapped significantly with the introns identified using microarrays ( p values of 6×10−47 and 8×10−7 for SPBC18H10 . 07 and SPAC30D11 . 14c , respectively ) . Moreover , this increase was accompanied by an accumulation of reads mapping to the corresponding exon-intron junctions , indicating that the mutations cause inefficient splicing of pre-mRNAs ( Figure S2B–E ) . Another gene , SPAC23A1 . 09 , accumulated 15 introns ( but not the corresponding exons ) , which did not overlap with those increased in other mutants ( Table S6 ) . SPAC23A1 . 09 encodes the fission yeast ortholog of Y14 ( also known as RNA-binding protein 8 ) , which is a component of the exon junction complex ( EJC ) , a multiprotein complex that is deposited upstream of exon-exon junctions and that has functions in RNA export , translational control and nonsense-mediated decay ( NMD ) [55] . Although the components of the EJC have not been studied in fission yeast , the EJC does not appear to have a function in NMD [56] . Our results suggest that the EJC may have a role in enhancing splicing efficiency for a subset of pre-mRNAs . Mutations in pab2Δ lead to the increased expression of both intronic and exonic regions of 21 genes [57] . In-depth biochemical analysis of this phenomenon using the rpl3002 transcript as a model revealed that Pab2 and the nuclear exosome are part of a polyadenylation-dependent pre-mRNA degradation pathway . We identified 4 genes with increased intronic expression in pab2Δ ( Table S6 ) , and 2 of these were also up-regulated at the exonic level . The only gene in common with the previous study was rpl3002 . The reason for this discrepancy is unclear . It is possible that our data processing was more stringent , as our lists have been derived from a larger number of biological repeats . On the other hand , microarrays do not have complete coverage of all introns and are also less sensitive . In the case of Red1 , which cooperates with Pab2 in nuclear RNA degradation , we found 8 introns overexpressed in red1Δ cells ( Figure 3C ) . Interestingly , the group of 8 introns contained 3 of the 4 introns up-regulated in pab2Δ cells ( Figure 3D ) . By contrast with usp102 and SPBC18H10 . 07 , the overexpression of 6 introns in red1Δ cells was accompanied by increased levels of the corresponding exonic probes ( Figure 3C ) . These observations suggest that Red1 has a role in the Pab2-dependent pre-mRNA decay pathway . Mutants in other potential regulators of RNA decay identified in this screen did not show concurrent overexpression of introns and exons , indicating that they regulate the degradation of mature mRNAs . Finally , pab2Δ and red1Δ mutants showed reduced expression of 21 and 18 introns , respectively , that showed a significant overlap and was not accompanied by a decrease in the corresponding exonic sequences . Therefore , Pab2 and Red1 appear to be required for the efficient splicing of a group of genes . The RBPs analysed in this work may affect transcription , either directly or indirectly through changes in RNA levels of transcription factor genes . Therefore , an important question is whether the observed changes in expression are due to alterations in transcription or RNA decay rates . To address this issue we measured RNA decay rates at the genome-wide level for nine selected strains that showed clear changes in RNA levels . We selected a protein known to regulate chromatin structure ( Set1 , which contains a single RRM ) and eight additional proteins ( Red1 , Pab2 , Zfs1 , SPAC17H9 . 04c , Rpm1 , Csx1 , Scw1 , and Rnc1 ) . To measure RNA stabilities we used an approach based on the in vivo labelling of RNAs with the modified nucleoside 4-thiouridine ( 4sU ) [8] , [12] . Cells are incubated with 4sU , which is incorporated into newly synthesised RNA at a rate determined by the stability of the RNA . As long as the system is under steady-state conditions ( that is , if RNA levels do not change over time ) , the fraction of labelled mRNA for a given gene can be used to estimate its decay rate . To measure this fraction , total RNA is prepared and 4sU–labelled RNA is specifically biotinylated . The biotinylated RNA can then be purified using streptavidin magnetic beads , and is compared to total RNA using two-colour custom DNA microarrays . We have previously set up this approach for S . pombe and showed that the steady state assumption holds for the conditions we employ [12] . As part of these experiments we measured RNA stabilities for 7 independent wild type samples . Reproducibility was high among biological replicates , with an average coefficient of variation of 12 . 6% . This high quality dataset improves and refines our previous estimates of half-lives ( Table S7 ) . Moreover , we provide the first quantification of stabilities of a subset of ncRNAs and introns in S . pombe ( Table S7 ) . Both ncRNAs and coding sequences showed similar half-lives ( medians of 30 . 5 and 31 minutes , respectively ) , while introns were shorter-lived ( 14 . 5 minutes ) ( Figure 4A ) . As expected , genes affected in set1Δ mutants did not show changes in stability ( Figure 4B and 4C ) . By contrast , Zfs1 targets predicted from the expression analysis were highly enriched among those mRNAs that displayed increased half-lives in the zfs1Δ mutant ( Figure 4D ) . Red1 and Pab1 are part of a system that down-regulates the expression of a group of meiotic mRNAs in vegetative cells , which are strongly overexpressed in red1Δ and pab2Δ mutants . We observed highly significant overlaps between mRNAs with increased stabilities in the mutants and those overexpressed , confirming and extending the observations that these proteins promote the decay of their mRNA targets ( Figure S3A and S3B ) . Moreover , ncRNAs induced in red1 mutants ( see above ) also displayed extended half-lives , indicating that Red1 promotes the decay of multiple ncRNAs . The effect of SPAC17H9 . 04c , which controls a regulon of genes involved in ribosomal synthesis , was also clearly posttranscriptional , as genes overexpressed in the mutant also showed enhanced stability ( Figure 4E ) . Similarly , the under-expression of genes encoding mitochondrial components in rpm1 mutants was strongly correlated with decreased stability ( Figure 4F ) . Finally , scw1 mutants showed correlated changes in mRNA stability and mRNA levels ( Figure 4G ) . By contrast , we did not observe stability changes for two RBP mutants , rnc1 and csx1 ( Figure S3C and S3D ) . Csx1 regulates RNA stability in response to oxidative stress [58] , while Rnc1 stabilises at least one mRNA in response to osmotic stress [59] . The reasons for this lack of correlation are unclear . It is possible that our system is not sensitive enough to detect very small changes in decay rates , and the changes in RNA levels in both of these mutants were subtle . Consistent with the complexity of the experiment to measure decay rates , it is our experience that this method is less sensitive than direct measurement of RNA levels . Alternatively , it is possible that Csx1 and Rnc1 only regulate RNA stability under conditions of stress . Recent studies have shown that changes in RNA stability may be compensated by alterations in transcription rates [31] . To investigate if this phenomenon is prevalent in our system we measured RNA stability for three RBP mutants that did not display changes in mRNA levels: SPAC25G10 . 01 , SPAC16E8 . 06c ( nop12 ) , and SPAC683 . 02c . In all three cases the fraction of mRNAs with altered RNA stability was between 0 and 0 . 04% , indicating that compensatory changes do not play an important role in these three strains . If a set of mRNAs is coregulated by a given RBP , it would be expected that the binding site of the RBP would be enriched in their sequences . Therefore , we searched for over-represented sequence elements in 5′-UTRs , 3′-UTRs and coding regions using software specifically designed for this purpose ( REFINE , [60] ) . We identified potential regulatory sequences in 12 regions corresponding to 8 proteins ( Table S8 and Figure 5 ) . Eight of the predicted binding sites were located in the 3′-UTRs , while 3 were identified in coding regions and one in 5′-UTRs . Red1 potential targets were enriched in sequences related to the Mmi1-binding site [61] in both their coding regions and 3′-UTRs . Similarly , we found that genes up-regulated in zfs1 mutants contained potential regulatory elements related to the published Zfs1 recognition site in both their 5′- and 3′-UTRs [39] . By contrast , mRNAs affected in SPAC17H9 . 04cΔ mutants were enriched in different sequence elements in their 3′-UTRs and coding regions . Surprisingly , the latter showed a very strong positional effect , with ∼67% of the sites located at positions 16 or 43 within the coding sequence . Finally , other RBPs that regulate groups of mRNAs with shared properties ( Rpm1 , Cip2 and Scw1 ) also contained enriched sequence elements ( Figure 5 ) . We chose the red1-enriched motif for further functional validation . A reporter construct [61] containing 8 tandem copies of the red1 potential regulatory motif ( UUAAAC ) in its 3′ UTR was not detectable by qPCR , while one with a mutated motif ( GUAAAC ) was clearly expressed ( Figure 6 ) . Deletion of red1 caused the UUAAAC reporter to be expressed at levels very similar to those of the reporter containing the mutated sequence ( Figure 6 ) , demonstrating that Red1 regulates mRNAs that contain the motif identified in our bioinformatics analyses . Inactivation of an RBP can cause changes in the stability of its targets , but may also affect other genes in an indirect manner . To investigate this question we used RIp-chip ( Ribonucleoprotein Immunoprecipitation analysed with DNA chips ) to identify RNAs bound to three RBPs identified in this screen: Red1 , Zfs1 , and Scw1 . The three proteins were epitope-tagged , immunopurified , and the associated RNAs identified using DNA microarrays ( Table S9 ) . In all three cases the RNAs identified by RIp-chip displayed highly significant overlaps with those RNAs overexpressed in the corresponding RBP mutants ( Figure S4 ) . By contrast , there was no enrichment in mRNAs under-expressed in the mutants , strongly suggesting that the three proteins destabilize the RNAs they bind to , and that decreases in gene expression associated with the mutations are indirect ( Figure S4 ) . Surprisingly , a substantial number ( ∼50 ) of RNAs bound by Zfs1 and Scw1 overlapped ( Table S9 ) . However , these RNAs were not enriched in those associated with Red1 or other any S . pombe proteins that we had previously investigated by RIp-chip [12] , [62]–[66] . Given the specificity of this overlap , it seems unlikely to represent a technical artefact , and could be explained if the two proteins are located to the same subcellular structure that co-purifies with them . It is possible that some RBPs do not have a role in vegetative cells and only function in specific developmental or environmental situations . For example , we have previously shown that the RBP Meu5 regulates mRNA stability during meiosis [12] . In vegetative cells , however , Meu5 is not expressed , and deletion of meu5 did not cause any changes in the transcriptome . To investigate if some of these RBPs have functions in specialised situations we monitored the ability of the RBP mutants to grow in 11 different conditions: low and high temperature , using non-fermentable carbon sources ( galactose ) , in the presence of hydroxyurea ( an inhibitor of DNA synthesis ) , calcofluor ( which impairs cell wall formation ) , cycloheximide ( an inhibitor of translation ) , methyl benzimidazol-2-yl-carbamate ( MBC , a microtubule poison ) , caffeine ( which overrides the S/M checkpoint ) , H2O2 ( that causes oxidative stress ) , cadmium ( a heavy metal ) , methyl methanesulfonate ( MMS , which induces DNA damage ) and high concentrations of KCl ( to trigger osmotic stress ) . All phenotypes were assessed by drop assays ( Figure S5 and Table S10 ) . In addition , we investigated the ability of the strains to undergo sexual differentiation ( mating and spore formation , Figure S6 and Table S10 ) . As a control we used a strain carrying a deletion in the caf1 gene , which encodes one of the catalytic subunits of the Ccr4-Not complex , and which has been reported to be sensitive to several stresses [67] , [68] . As expected , caf1Δ cells exhibited phenotypes in 9 different conditions ( Table S10 ) . 39 strains ( 53% of the collection ) showed at least one phenotype ( Figure S5 ) . However , only nine displayed three phenotypes or more , and the four strains that showed most phenotypes were affected in genes involved in general expression pathways such as mRNA deadenylation ( caf1 ) , splicing ( usp102 ) , translation initiation ( sce3 ) and nucleosome remodelling ( spt6 ) . These data indicate that the screen is highly specific . The most common phenotype was defects in sexual differentiation , which affected 16 strains ( 22% ) . Sensitivity to cadmium , caffeine and cycloheximide were also common ( ∼15% ) , the latter suggesting that some of these RBPs may have functions in translation . Interestingly , we found three strains that were strongly resistant to cadmium . The sensitivity to other environmental stresses was highly specific . For example , only three strains were sensitive to oxidative stress ( caf1 , rpm1 and csx1 ) , and a single one was unable to grow using a non-fermentable carbon source ( rpm1 ) . Altogether , our results demonstrate that many of the RBPs studied here have functions in the response to specific stress conditions and during cellular development .
We have systematically characterised the vegetative transcriptome and the function of the majority of genes encoding sequence-specific RBPs in S . pombe . Our results offer new biological insights and provide a valuable resource for the study of posttranscriptional control in fission yeast . The characterisation of some previously studied genes provides new clues about their function . For example , we show that Red1 participates in a pre-mRNA degradation pathway , probably in cooperation with Pab2 . Moreover , we have found that Red1 regulates the expression of long ncRNAs through the control of their stability . We have also identified new pathways ( Pan2/Pan3 ) involved in the regulation of ncRNA expression . Our systematic study has identified and characterised RBPs that regulate mRNA decay in S . pombe cells and has provided lists of their potential targets . The RBP putative targets presented common features , such as being co-expressed in response to stress ( mug187 ) or during meiosis ( pab2 and red1 ) , encoding proteins with similar localizations ( SPAC17H9 . 04c and rpm1 ) , or coding for proteins with related functions ( scw1 , SPAC17H9 . 04c and rpm1 ) . This is consistent with the concept of the posttranscriptional operon or regulon , which states that RBPs coordinate the expression of genes with related functions at the posttranscriptional level [69] , [70] . Perhaps the most surprising result from this work is the fact that only a relatively small fraction of genes ( 16 . 2% ) is affected in the whole RBP deletion collection , suggesting that the function of non-essential RBPs is not sufficient to explain the large range of mRNA half-lives in fission yeast . There are several non-mutually exclusive explanations can account for this phenomenon . First , genetic redundancy may provide backup pathways that compensate for the lack of single RBPs . Second , it is possible that the majority of half-lives are determined by essential RBPs , which were not analysed in this study . Third , the general decay machinery might be able to interact with specifically with mRNAs in a gene-specific way independently of RBPs . Fourth , recent studies have revealed the existence of widespread connections between transcription and RNA degradation ( reviewed in [31] ) , in which the effects of mutations affecting RNA decay is compensated by changes in transcription . However , we did not find any indication of this phenomenon in the three strains that we examined . Finally , promoter sequences can regulate RNA stability without involvement of cis elements on the mRNAs , presumably through the cotranslational recruitment of proteins to general mRNA features ( such as the cap or the poly ( A ) tail ) [71] , [72] . Our work provides a framework that will allow the examination of these hypotheses in fission yeast and other organisms .
Standard methods and media were used [73] . For transcriptome analysis cells were grown in Edinburgh Minimal Medium ( EMM ) at 32°C to a cell density of 8×106 cells/ml . For spot assays cells were grown in yeast extract medium ( YE ) to a concentration of 8×106 cells/ml , and plated in 10-fold dilutions . The following concentrations were used in YE agar plates: H2O2 at 1 , 1 . 5 and 2 mM; CdSO4 at 0 . 3 , 0 . 4 , 0 . 5 and 0 . 6 mM; hydroxyurea at 5 and 10 mM; methyl benzimidazol-2-yl-carbamate ( MBC ) at 5 µg/ml; cycloheximide at 10 , 20 and 30 µg/ml; methyl methanesulfonate ( MMS ) at 0 . 0027%; calcofluor at 0 . 5 and 1 mg/ml; caffeine at 5 , 10 and 15 mM; KCl at 0 . 8 and 1 M . For growth in non-fermentable sources , glucose in YE was replaced with 2% galactose and 0 . 1% glucose . For assessment of mating and sporulation cells were grown on malt extract agar ( MEA ) . All plates were incubated at 32°C . For testing temperature-sensitive growth , cells were plated on YE agar and incubated at either 20°C or 36°C . Table S11 lists all the strains used in this work . Cells were made homothallic ( h90 ) and auxotrophic markers were removed by crossing before the experiments . The presence of a correct deletion was first assessed by examining the microarray signal for probes corresponding to the deleted gene . When this approach produced ambiguous results ( for example , if a gene was expressed at low levels ) , we performed gene-specific diagnostic PCR . Overall , we verified the deletion for 74 strains and confirmed that the expected gene was not deleted in 12 others ( Table S1 ) . Total RNA was purified using phenol extraction [74] . Fluorescently labelled cDNA was prepared from total RNA using the SuperScript Plus Direct cDNA Labelling System ( Life Technologies ) as described by the manufacturer , except for the following modifications: 8 µg of total RNA was labelled in a reaction volume of 15 µl . 0 . 5 µl of 10× nucleotide mix with labelled nucleotide were used ( 1/3 of the recommended amount ) and 1 µl of a home-made dNTP mix ( 0 . 5 mM dATP , 0 . 5 mM dCTP , 0 . 5 mM dGTP , 0 . 3 mM dTTP ) was added to the reaction . All other components were used at the recommended concentrations . Note that these changes are essential to prevent dye-specific biases . Labelled cDNAs were hybridised to oligonucleotide microarrays manufactured by Agilent as described [63] . Microarrays were scanned with a GenePix 4000A microarray scanner and analysed with GenePix Pro 5 . 0 ( Molecular Devices ) . mRNA decay rates were determined using in vivo labelling with 4-thiouridine ( 4sU ) [12] . Briefly , cells were grown in EMM at 32°C , and mRNAs were labelled by the addition of 4sU to the medium at a final concentration of 75 µg/ml . Cells were collected after incubation with 4sU for 7 or 10 minutes depending on the strain . An isogenic wild type was processed in parallel with each of the mutants . Total RNA was phenol-extracted and 4sU-labelled RNA was biotinylated and purified as described [12] . Finally , 4sU-labelled fractions and total RNA were compared using DNA microarrays as described above . Immunoprecipitation of TAP-tagged proteins was carried out using monoclonal antibodies against protein A ( Sigma ) , and myc-tagged proteins were purified using the 9E11 monoclonal antibody ( Abcam ) . For Scw1-TAP and Zfs1-TAP RIp-chips were performed as described [62] except for the following modifications: 1 ) the lysis buffer contained 1 mM PMSF and 1∶100 protease inhibitor cocktail ( sigma P8340 ) and 2 ) magnetic beads containing the immunoprecipitate were resuspended in 50 µl of wash buffer containing 1 mM DTT , 1 unit/ml of SuperaseIN ( Ambion ) and 30 units/ml of AcTev protease ( Life Technologies ) . The solution with the beads was incubated for 1 h at 19°C , the supernatant recovered and RNA extracted using PureLink RNA micro columns ( Life Technologies ) according to the manufacturer's instructions . The RNA was eluted from the column in 12 µl and used for labelling without amplification . For Red1-myc we followed a published protocol [62] , except that the lysate was prepared in the following buffer: 10 mM Hepes pH 7 . 4 , 100 mM KCl , 5 mM MgCl2 , 25 mM EDTA , 0 . 5% NP-40 , 1% Triton X-100 , 0 . 1% SDS and 10% glycerol containing 1 mM PMSF and 1∶100 protease inhibitor cocktail ( sigma P8340 ) . Microarray data for transcriptome analysis were normalized using Loess , and for RNA stability determination expression ratios were median-centred . Differentially expressed genes were identified using Significance Analysis of Microarrays ( SAM ) [37] . Significance of the overlap between gene sets was determined using Fisher's exact test . Comparison with published microarray datasets was performed as follows: For pab2 we used a list of 31 significantly up-regulated genes ( Table 1 in reference [41] ) , for red1 we used Tables S2 and S3 , which report lists of 121 and 30 genes that are up-regulated or down-regulated , respectively , at least two-fold [42] . For zfs1 we applied SAM to data from four microarrays from wild type cells and from zfs1 mutants [39] using an FDR of 0 . For the comparison with ncRNAs from red1Δ cells we used Table S2 , which contains 269 ncRNAs [44] . The analysis of RIp-chip experiments was performed as described [12] . Briefly , all RNAs present in the immunoprecipitate were ranked by their enrichment levels ( ‘physical targets’ ) . We then made lists of physical targets containing increasing amounts of genes ( starting with the most enriched ) , and quantified the number of genes in each of the lists whose expression was affected in the corresponding RBP mutant . As a control , the analysis was repeated with a randomized list of physical targets . The initial gradient of the RIp-chip data is higher than that of the randomized data , indicating that the physical targets are enriched in genes affected by the mutation . When the gradient of both curves converges , the RIp-chip data stop having predictive value for the identification of regulated mRNAs . This point was chosen as a cut-off for the definition of RBP targets . RNAs identified in both biological repeats were selected . Significance of the overlap between gene sets was determined using Fisher's exact test . We searched for enriched sequence elements in 5′- , 3′-UTRs and coding sequences using REFINE [60] with default parameters . The searches used sequence databases of UTRs and coding regions generated using information from GeneDB ( http://old . genedb . org/ ) , now PomBase ( http://www . pombase . org/ ) , on May 9 , 2011 [36] . A threshold E value of 10−8 was used for selection of the motifs displayed in Table S8 . Sequence logos were generated with WebLogo ( http://weblogo . berkeley . edu/ ) [75] . Wild type leu1-32 or red1Δ::kanMX6 leu1-32 cells were transformed with plasmids pRGT1-GFP-TTAAAC8x or pRGT1-GTAAAC8x [61] . Both plasmids express GFP under the control of the adh1 promoter and contain 8 copies of the indicated sequence motifs . Cells were grown in EMM to a concentration of 107 cells/ml . Total RNA was extracted using a hot phenol protocol [74] . 20 µg of total RNA were treated with 2 units of Turbo DNAse ( Life Technologies ) and purified using a PureLink RNA Micro kit ( Life Technologies ) following the manufacturer's protocol . 1 µg of purified RNA per sample was reverse-transcribed using Superscript III ( Life Technologies ) . Quantitative analysis of RNA levels was performed using Sybr Green JumpStart Taq ReadyMix ( Sigma ) in a real-time PCR machine ( Rotor-Q Gene , Quiagen ) using the following program: 10 minutes at 95°C , then 40 cycles of 95°C for 10 seconds , 60°C for 15 seconds and 72°C for 30 seconds , with a final 5-second melting ramp of 1°C steps ( from 50°C to 95°C ) for acquisition . The following primers were used: GFP-F ( CATCATGGCAGACAAACAAAA ) , GFP_R ( AAAGGGCAGATTGTGTGGAC ) , ACT2_F ( CCGGACTCGAGAAGAAACAT ) and ACT2_R ( AACCACCTTTTTCCGCTCTT ) . Quantification of relative levels was performed as follows: Ratio ( GFP/actin ) = ( Eff∶GFP-Ct∶GFP ) / ( Eff∶actin-Ct∶actin ) , where Eff∶GFP and Eff∶actin represent the amplification efficiencies , and Ct∶GFP and Ct∶actin are the critical cycles . A total of 219 microarray experiments were performed . All microarray and sequencing data have been deposited in ArrayExpress with accession numbers E-MTAB-2314 ( microarray expression experiments ) , E-MTAB-2317 , E-MTAB-2318 and E-MTAB-2712 ( stability data ) , E-MTAB-2709 ( RIp-chip experiments ) and RNA-seq of splicing mutants ( E-MTAB-2695 ) . | Messenger RNAs ( mRNAs ) are the molecules that relay the information from genes ( DNA ) to proteins . Cells contain different amounts of each mRNA type depending on their function and their situation . The quantity of each mRNA depends on the balance between its production ( transcription ) and its degradation ( mRNA decay ) . Recent studies have shown that the rate at which each mRNA is degraded is specific for every gene , but little is known about how this is regulated . In this work , we look at the role of a class of proteins that bind to RNA molecules ( RNA-binding proteins , or RBPs ) in the regulation of RNA decay . By systematically examining cells in which a single RBP has been inactivated we identify those that are important for RNA degradation . We found RBPs that make mRNAs more stable ( that is , they are degraded more slowly ) and others that make them unstable . These RBPs control the RNAs of genes with common features , suggesting that they provide a way of coordinating the function of groups of genes . However , for many genes we did not find RBPs that control their stability , indicating that other players are important to regulate RNA degradation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"organisms",
"biochemistry",
"rna-binding",
"proteins",
"rna",
"fungi",
"proteins",
"gene",
"regulatory",
"networks",
"genetics",
"schizosaccharomyces",
"pombe",
"biology",
"and",
"life",
"sciences",
"molecular",
"genetics",
"computational",
"biology",
"schizosaccharomyces",
"rna",
"stability",
"yeast"
]
| 2014 | Systematic Analysis of the Role of RNA-Binding Proteins in the Regulation of RNA Stability |
Multilevel selection has been indicated as an essential factor for the evolution of complexity in interacting RNA-like replicator systems . There are two types of multilevel selection mechanisms: implicit and explicit . For implicit multilevel selection , spatial self-organization of replicator populations has been suggested , which leads to higher level selection among emergent mesoscopic spatial patterns ( traveling waves ) . For explicit multilevel selection , compartmentalization of replicators by vesicles has been suggested , which leads to higher level evolutionary dynamics among explicitly imposed mesoscopic entities ( protocells ) . Historically , these mechanisms have been given separate consideration for the interests on its own . Here , we make a direct comparison between spatial self-organization and compartmentalization in simulated RNA-like replicator systems . Firstly , we show that both mechanisms achieve the macroscopic stability of a replicator system through the evolutionary dynamics on mesoscopic entities that counteract that of microscopic entities . Secondly , we show that a striking difference exists between the two mechanisms regarding their possible influence on the long-term evolutionary dynamics , which happens under an emergent trade-off situation arising from the multilevel selection . The difference is explained in terms of the difference in the stability between self-organized mesoscopic entities and externally imposed mesoscopic entities . Thirdly , we show that a sharp transition happens in the long-term evolutionary dynamics of the compartmentalized system as a function of replicator mutation rate . Fourthly , the results imply that spatial self-organization can allow the evolution of stable folding in parasitic replicators without any specific functionality in the folding itself . Finally , the results are discussed in relation to the experimental synthesis of chemical Darwinian systems and to the multilevel selection theory of evolutionary biology in general . To conclude , novel evolutionary directions can emerge through interactions between the evolutionary dynamics on multiple levels of organization . Different multilevel selection mechanisms can produce a difference in the long-term evolutionary trend of identical microscopic entities .
Consideration of selection acting on multiple levels of biotic organization is important for understanding of biological evolution in general [1]–[6] . In the studies of prebiotic evolution , it has been shown that some form of multilevel selection is even necessary for the maintenance and evolution of complexity in the interacting RNA replicator system [7]–[12] . The RNA-like replicator system is considered one of the simplest chemical systems that can undergo Darwinian evolution in a self-sustained manner [13] , [14] . Hence , RNA replicators have been suggested as the central player of prebiotic evolution in the RNA world hypothesis [15]–[18] . Besides whether or not such replicators—or its analogues [19]—can actually exist ( see [20]–[25] , for recent progress on this ) , an interesting question is whether such a chemical replicator system can increase its complexity through evolution and approach the biotic system as we know it . The importance of multilevel selection in prebiotic evolution is based on two problems that arise in the evolution of replicator systems . Firstly , there is a fundamental problem about the accumulation of information in exponentially growing replicators . That is , the maximal length of sequence patterns that can be maintained under the mutation-selection process in a single replicator quasi-species is severely limited by high mutation rates , which are supposed in primordial replication processes based on RNA molecules [13] , [26] ( see also Text S1 note 1 ) . Hence , a solution to information accumulation has been sought in the symbiosis of multiple species of replicators . As such a solution , a classical study suggested the so-called hypercycle , in which every replicator species catalyzes replication of another species , forming a cyclic network of cooperative interactions [27] . However , there is an inherent problem in such a cooperatively interacting replicator system , in that selection acting on the level of individual replicators favors only the evolution of better templates , but does not favor—even disfavor [28]—the evolution of better catalysts [7] , [8] . To overcome this problem , spatial structure in a population has been suggested . Generally speaking , spatial population structure can be classified according to whether it is implicit or explicit: Implicit population structure arises from the birth-death-migration process of individuals themselves , whereas explicit population structure is imposed to a population by some external boundaries ( of course , the external factors can depend on the activity of individuals ) . In the context of prebiotic evolution , both kinds of population structure have been investigated: ( i ) spatial self-organization of populations in a diffusion-limited , surface-bound replicator system as implicit structure [4]; ( ii ) compartmentalization of replicators by vesicles as explicit structure [9] , [29] . Historically , the two kinds of population structure were given separate consideration for the interests on its own . In both cases , however , the essential process that makes the difference from a well-mixed system is the formation of multilevel selection , i . e . selection operating on the level of microscopic entities ( i . e . individual replicators ) and on the ( various ) levels of mesoscopic entities , such as ( spiral ) waves and protocells , that arise from spatial population structuring . In a previous study , we constructed a computational model that can simulate a surface-bound replicator system and compartmentalized replicator system in a unified framework . We therewith investigated the two different types of spatial population structure—explicit versus implicit—with respect to their influence on the macroscopic stability of different evolving replicator systems [30] . The current study aims to extend the previous study: It makes a direct comparison between the spatial self-organization and compartmentalization with respect to their effects on the eco-evolutionary dynamics of a simple interacting replicator system , particularly , by focusing on the interactions between the dynamics of microscopic and mesoscopic entities . For this sake , we adopt the following research strategies . Firstly , we consider an identical model of interacting replicators in the surface-bound system and in the compartmentalized system so that , while the two systems differ in mesoscopic entities , they share identical microscopic entities . Secondly , we design vesicle-level processes such that they do not introduce an extra burden for survival that is independent of the replicator dynamics itself; e . g . , we neglect the problem of substrate uptake through membranes ( see [31] , [32] for experimental studies on this issue ) . These considerations allow us to focus on the effect of spatial population structure itself . Thereby we study how different types of spatial population structure achieve the macroscopic stability of a replicator system , and how the dynamics of different mesoscopic entities influence the evolutionary dynamics of microscopic entities and vice versa .
The replicator model investigated here consists of two types of molecules: replicase and parasite . The replicase can catalyze replication of other molecules , whereas the parasite cannot . The parasite switches between two conformations , viz . folded state and template state . When a parasite is in the folded state , it can facilitate the growth of the vesicle in which it resides ( explained later ) , but cannot be replicated by the replicase; when in the template state , it can be replicated , but cannot facilitate the vesicle growth . We assume the conformation switching is so fast that it is always in equilibrium ( see also Text S1 note 2 ) . Hence , the concentration of parasites in the folded state ( ) and that in the template state ( ) can be expressed as and respectively , where is the total concentration , and , and is the equilibrium constant of . Thus , the current replicator system can be represented as follows: ( 1 ) where and denote a replicase and parasite molecule respectively; and denotes a complex molecule between and and that between and respectively; represents the generalized resource for replication . In Reaction 1: The consideration of complex formation is to take into account the fact that replication takes finite amount of time , during which the replicase cannot be replicated . Complex formation thus considerably disadvantages the replicase over the parasite ( see [28] for more details; see also [33] ) . The replicator dynamics was modeled in the framework of stochastic cellular automata ( CA ) . The model consists of a two-dimensional grid of squares , where one square can contain at most one molecule . Empty squares are considered to represent the generalized resource ( ) , which limits the maximum number of molecules the system can sustain globally and locally . The model's dynamics are run by randomly choosing one grid square and then locally applying the algorithm that simulates Reaction 1 and diffusion ( both prohibited to occur across grid boundaries ) . The reaction and diffusion algorithm employed here simulates the chemical reaction dynamics in such a way that the result approaches that of the Gillespie algorithm in the limit of with being constant , where denotes the rate of diffusion . To investigate the evolution of replicators , we introduced “mutations” in and . A newly produced parasite inherits the values of and from its template , but mutation can modify either or by adding to it a number uniformly distributed in . Moreover , and are bounded in ( see also Text S1 note 3 ) . The mutation of and occur with a probability and respectively , and they are mutually exclusive . Vesicle-level processes were modeled by using the so-called Cellular Potts Model ( CPM ) [34] , [35] . The CPM is a two-scale stochastic CA: It explicitly defines a mesoscopic entity , “vesicle” , by a set of grid squares having an identical state; at microscopic scale , the updating rules tend to minimize the interface between different vesicles , while at mesoscopic scale the updating rules tend to keep the volume of a vesicle ( i . e . the number of the squares constituting a vesicle ) at the target volume . We implemented a two-dimensional CPM of squares . We then superimposed it onto the replicator CA plane and coupled the dynamics of the two as follows . Firstly , the molecules in the replicator CA are forbidden to permeate though the vesicle boundaries in the CPM ( i . e . , replicators cannot diffuse across vesicle boundaries , and vesicle boundaries cannot move over molecules ) . Secondly , the dynamics of the target volume of vesicles are governed by the increase due to the occurrence of Reaction 1 ( c ) and the decrease due to spontaneous decay . That is , if Reaction 1 ( c ) happens inside a vesicle , its target volume is increased by 1 ( if it happens outside vesicles , it is ignored ) ; hence , Reaction 1 ( c ) can be considered to represent membrane production . Moreover , the target volume decays with rate ( the decay rate of target volume ) . It is worth noting that plays a crucial role for the competition among vesicles because the greater is , the greater chance a vesicle has to expand its actual volume and grow . Thirdly , a vesicle divides when its actual volume exceeds a threshold ( “reproduction” ) : A vesicle is divided along the line of the second principle component; the internal replicators are distributed ( or “inherited” ) between the two daughter vesicles according to the location ( if a complex is divided between the two vesicles , it is dissociated ) ; and the target volume is also distributed between them in proportion to their volume after division . It is also worth noting that determines the size of vesicles and , thus , the number of replicators in vesicles , which is a crucial parameter of compartmentalization . Additionally , the death of vesicles is not simulated as an explicit process , for it happens implicitly through the dynamics already specified ( see also Discussion ) . Importantly , our compartmentalized replicator model ignores the transport process across vesicle boundaries and the resource for the target volume growth . This simplification is to avoid introducing an extra constraint for survival that is not considered by the replicator model per se ( cf . the surface model ) , which allows us to directly compare spatial self-organization and compartmentalization with respect to their effects on the replicator dynamics . Also notable is that vesicle-level selection is nearly at optimal efficiency , for a difference in , which is unbounded , is always reflected in the competence in volume expansion .
The dynamics of the replicator system without population sturcutre can be considered the point of reference for the dynamics with population structure . A simple ordinary differential equation ( ODE ) model was constructed that describes the well-mixed system of one replicase and one parasite species according to Reaction 1: ( 2 ) where , , and denote the concentration of R , L , and respectively; and the dots denote time derivative; and . We study the behavior of Eqn . 2 as a function of and since we later investigate the evolution of these parameters in the systems with spatial population structure . We numerically calculated the equilibrium of Eqn 2 as shown in Fig . 1 . The result shows that the values of that allow the stable coexistence of and shift to higher values as increases . The result can be explained with ease: Increasing gives an advantage to the parasite for replication; hence , in order to allow the coexistence , this must be compensated by increasing , which reduces the fraction of time the parasite spends as template . From this argument , one can also expect that if a parasite species with a greater and/or smaller value is introduced to the system , it will out-compete the original parasite; i . e . , there is selection pressure to increase and to decrease . Indeed , this was numerically confirmed by extending Eqn . 2 to include another parasite species ( data not shown; the extended version of Eqn . 2 is shown in Text S1 ) . Therefore , if the evolution of and/or is allowed , a well-mixed system will eventually go extinct due to the evolution of too harmful parasites . This was also confirmed by the CA model simulating a well-mixed system ( data not shown ) . In summary , the well-mixed replicator system is evolutionarily unstable , so that some sort of spatial population structure is necessary for the feasibility of the evolving interacting replicator system ( see [12] on this point discussed with the model explicitly considering the genotype-phenotype-interaction mapping of replicators; see [28] for more detailed analysis on a similar ODE model ) . In this section , we examine the evolutionary dynamics of the replicator system with the two types of spatial population structure , viz . compartmentalization and spatial self-organization . We will examine whether the replicator system can survive despite the evolution of parasites and what kind of evolutionary dynamics the system will display . The surface model and compartment model were initialized by inoculating the system with small populations of the replicase and parasite of an equal size . and were allowed to mutate ( the initial population was homogeneous ) , while other parameters were fixed . In the compartment model , molecules were randomly placed inside one large vesicle . In the surface model , replicase molecules were placed in a half circle , and parasite molecules were placed in the other half circle . In the compartment model , the value of ( diffusion ) was set so great as to remove the effect of spatial self-organization within vesicles in order to simplify the comparison with the surface model ( see also Text S1 note 4 ) ; otherwise , the two models had identical values in the shared parameters . To obtain the visual image of our models , snapshots of the simulations are shown in Fig . 2 . Moreover , Videos S1 and Video S2 depict the spatio-temporal dynamics of the compartment model and that of the surface model respectively ( for visibility , Video S1 depicts a smaller scale simulation than that shown in Fig . 2 ) . As Video S2 ( surface model ) shows , mesoscopic patterns—namely , traveling waves—emerge through the spatial self-organization of the replicator population , which contrasts with the compartment model where mesoscopic patterns—i . e . vesicles—were externally imposed ( see [28] for more description on the spatio-temporal dynamics of such waves ) . In the compartment model , a vesicle expands its volume as the internal replicators multiply , extending into an empty area or pushing other vesicles away ( i . e . inter-vesicle competition ) , and it divides when its volume exceeds the threshold . Once a while , a vesicle also shrinks—or gets squeezed—and disappears from the system ( i . e . dies ) in concurrence of the extinction of the internal replicators . The long-term behavior of the simulations is depicted as the evolutionary trajectories of the population average of and in Fig . 3 ( black and red lines ) . The trajectories can be separated into two phases: short-term evolution and long-term evolution . In the former , the trajectories go to a contour that gives a ( mathematical ) functional relationship between and , which indicates the emergence of a trade-off situation in parasites regarding the affinity towards the replicase ( ) and the availability of templates ( ) . In the latter , the trajectories go along the contour , increasing and . These results show that both the surface model and the compartment model allowed the stable coexistence of the replicase and parasite despite the evolutionary instability of the replicator system explained before . Moreover , the two models exhibited a qualitatively identical evolutionary trend such that the parasite , through evolution , increased the fraction of time it spent in the folded state while it also increased the affinity towards the replicase . This result is surprising , given that the folded state has no predefined functionality in the surface model ( in fact , it prevents the replication ) , whereas it does have a predefined functionality in the compartment model ( i . e . to facilitate the vesicle growth ) . To understand these results , we next delve into each model . In this section , we show that the population dynamics of traveling waves exhibit the property of multiplication , variation and inheritance , and therby it ensures the macroscopic stability of the replicator system . Moreover , we analyze what kind of selection pressure exists among waves , which turns out to be qualitatively different from the vesicle-level selection that arises by default . • The two multilevel selection models are quite similar in how they achieve the macroscopic stability of the replicator system: the evolutionary dynamics on the microscopic entities ( i . e . replicators ) are counteracted by the evolutionary dynamics on the mesoscopic entities ( i . e . vesicles or traveling waves ) . - In the compartment models , the vesicle-level selection operates on the variability in internal replicator systems generated by the stochastic evolutionary dynamics of replicators . - In the surface model , selection operates on the level of traveling waves , which have the feature of multiplication , variation and inheritance . • However , the two types of mesoscopic entities differ in their stability in isolation . - In the compartment model , a vesicle is an externally imposed mesoscopic entity ( explicit multilevel selection ) , and it is less persistent . - In the surface model , a traveling wave pattern is a self-organized mesoscopic entity ( implicit multilevel selection ) , and it is thus more persistent than a vesicle ( if it is too unstable , it would not self-organize ) . • The difference in the stability of mesoscopic entities results in the difference in the focus of mesoscopic selection . - The vesicle-level selection , by default , operates for the longevity of vesicles due to its greater instability . - The wave-level selection operates for the fecundity of waves ( i . e . the generation of new traveling waves ) . • Because multilevel selection keeps the evolution of too severe parasitism at bay , parasites have a trade-off situation between ( i . e . affinity to replicase ) and ( i . e . template availability ) . Under this trade-off , parasites can adopt two kinds of strategies in the association-dissociation reaction with replicase: ( A ) increasing the effective rate of association—smaller and —and ( B ) biasing the equilibrium towards association—greater and . Strategy A weakens the deterministic flow of the replicator dynamics while prohibiting the transient growth of a population consisting of a small number of replicases and parasites . Strategy B strengthens the deterministic flow while enhancing such transient growth . These strategies gain selective differences through the interactions between the dynamics of microscopic entities and those of mesoscopic entities . This produces a novel trend in the long-term evolution of the replicator system , which can differ between the two multilevel selection models . - In the compartment models , the death rate of vesicles depends on the stability of the coexistence between the replicase and parasite in the internal replicator system . If the coexistence is deterministically stable , strengthening the deterministic flow of the internal replicator dynamics is favored . If the coexistence is deterministically unstable , weakening the deterministic flow is favored . The evolutionary dynamics of internal replicator systems are fast when the mutation rate of replicators is high and the population size of internal replicator systems is large . In this case , the coexistence is likely to be deterministically unstable; therefore , weakening the deterministic flow of the replicator dynamics ( i . e . Strategy A ) is favored . ( Similarly , if the internal replicator evolutionary dynamics are slow , Strategy B is favored . ) However , this default direction of the vesicle-level selection can be overruled by an additional selection pressure arising from the ( predefined ) functionality of the folded state to facilitate the vesicle growth . - In the surface model , the establishment of new waves depends on the ( transient ) growth of a population consisting of a small number of replicases and parasites . Therefore , Strategy B is favored . In this section , we compare the surface model and the compartment model with respect to how the macroscopic stability responds to the change of either the diffusion rate ( ) in the surface model or the threshold volume for division ( ) in the compartment model , whereby we illustrate an interesting difference between the two models . Our choice to focus on and is based on two reasons . Firstly , previous studies have suggested that these parameters significantly affect the macroscopic stability of the replicator system [11] , [28]–[30] , [33] , [42] , [43] . Secondly , and play a similar role in how they limit the macroscopic stability; i . e . , increasing and increasing increase the number of replicators involved in local replicator systems that can be considered independent of each other due either to vesicle boundaries or to spatial distance ( one traveling wave can be considered to consist of multiple such local replicator systems as seen from the heterogeneity within a wave shown in Fig . 8 ) . Firstly , we investigated what might be called the “ecological” stability of the system , i . e . the range of and for which the replicator system exhibits the macroscopically stable coexistence of the replicase and parasite with no mutation ( i . e . ) . As seen from Fig . 11 , while the survival regions from the two models are similar in topology , they significantly differ in its response to the change of or . That is , the survival area in the surface model greatly varies as changes whereas that of the compartment model varies relatively little as changes ( the modified compartment model showed qualitatively the same result; see Fig . S1 ) . Additionally , we mention that the compartmentalized system was not viable for , which was most likely because of too great assortment load [29] , [42] . Also , the spatial pattern in the surface model appeared different for large diffusion rates as shown in Video S5 ( see Text S1 for more on this point ) . Secondly , we investigated the “evolutionary” stability of the system , i . e . the maximal tolerable value of ( ) for which the system exhibits the macroscopically stable coexistence of the replicase and parasite ( with and ) as shown in Fig . 12 ( the modified compartment model showed qualitatively the same result; see Fig . S2 ) . The result shows that significantly changes as the parameter varies in both models , and the curves of are very similar to each other . These results are in stark contrast to the response of the survival regions seen above . These results indicate that there are two different aspects in the macroscopic stability of replicator systems . One is the range of rate constants for which a system displays the macroscopic stability—“ecological stability” . The other is the degree of perturbation to rate constants ( i . e . mutation in the sense used here ) for which a system displays the macroscopic stability—“evolutionary stability” . Interestingly , these two aspects do not necessarily correspond to one another . Consequently , the two models showing different degrees of ecological stability can show similar degrees of evolutionary stability . Fig . 11 and 12 depict the general tendency that the macroscopic stability decreases as either or increases . This is explained by the fact that increasing the number of replicators involved in the local replicator systems reduces the stability of the whole system in two ways . Firstly , it decreases the stochasticity in the dynamics—both ecological and evolutionary—of the local replicator systems , which diminishes the relative impact of higher-level selection through reducing the variability among the local systems [9] ( in the surface model , the local variability referred here also includes the heterogeneity within each wave ) . Secondly , it increases the frequency at which stronger parasites appear through mutation in a local replicator system per unit time [29] , [44] , which speeds up the evolutionary dynamics of local replicator systems ( see also the caption to Fig . 6; note that this is irrelevant to the ecological stability by definition ) . It is worth mentioning that the expansion of the survival region by decreasing decreases as increases as shown in Fig . 11B and S1B ( see the survival range of as a function of ) . This is nicely explained by the first effect explained above and the fact that the deterministic flow of the replicator dynamics strengthens as increases . Additionally , we note that the fraction of replicases within vesicles increases as decreases . A similar observation has been made in other models [30] , [45] , [46] ( esp . [45] ) However , as we saw above , the survival area of the surface model is far more sensitive to than that of the compartment model is to ( Fig . 11 ) . This can be explained by another effect of changing ; that is , decreasing makes it more difficult for a parasite molecule to be in contact with replicase molecules than for a replicase molecule ( see [43] for more quantitative investigation on this ) . This means that decreasing directly disadvantages the parasite . Therefore , the survival area substantially changes as a function of . However , despite this additional effect , the dependency of of the surface model is not necessarily more sensitive to than that of the compartment model is to ( Fig . 12 ) . This is explained by the fact that the ecological stability is “offset” by the evolutionary dynamics; i . e . , for greater mutation rates the replicator system tends to go to the lower boundary of the survival region through evolution ( it is assumed that the ecologically stable region does not cover the entire span of the parameter range; see the case of in Fig . 12A ) . To summarize , the macroscopic stability of the interacting replicator system has two different aspects: ecological stability and evolutionary stability . In the surface model , decreasing substantially enhances the ecological stability because decreasing directly decreases the chance of parasites' being in contact with replicases . Contrastingly , in the compartment model , decreasing does not have such a direct effect; hence , its enhancement of the ecological stability is more moderate . However , this apparent advantage of the surface model with respect to the ecological stability is offset by the tendency of the evolutionary dynamics to bring the system to an edge of the ecologically stable parameter region , which indicates the greater importance of the evolutionary stability relative to the ecological stability . Consequently , the two models display a similar response in the evolutionary stability as a function of or .
The current study compared two multilevel selection models of replicator systems that had identical microscopic entities , but had qualitatively different mesoscopic entities . Despite the difference in the mesoscopic entities , we found that the two models were quite similar in how they achieved the macroscopic stability of the replicator system . Moreover , we also discovered an emergent trade-off situation in microscopic entities , which arose due to the multilevel selection ( we note that a similar trade-off situation was previously discovered by van Ballegooijen and Boerlijst [41] ) . Interestingly , under this trade-off situation , microscopic entities displayed novel long-term evolutionary trends , which originated from the interactions between the dynamics of microscopic entities and mesoscopic entities . Furthermore , in contrast to the similarity mentioned above , the two models could sharply differ in the direction of this evolutionary trend , which was explained in terms of the difference in the stability between self-organized mesoscopic entities and externally imposed mesoscopic entities . The surface model showed that the parasite , through long-term evolution , increased the time it spent in the state in which it could not function as template , despite the fact that no functionality was predefined for this state . Since the folding of an RNA molecule is likely to reduce the template activity , this result can be interpreted as the implication that in the diffusion-limited surface-bound system the parasite can evolve stable folding “for free” , i . e . without any specific functionality in the folding . The evolution of stable folding might be used as substrate for the further evolution of new functionality . Hence , the current study revealed a novel advantage of spatial self-organization for the evolution of complexity in RNA-like replicator systems . In the compartment model , we found a simple relationship between the persistency ( i . e . longevity ) of a vesicle and the dynamical property of the replicator system inside the vesicle . That is , if the evolutionary dynamics of internal replicator systems are fast , the coexistence of the replicase and parasite in internal replicator systems is deterministically unstable; hence , weakening the deterministic flow of the internal replicator dynamics would increase the longevity of vesicles . Similarly , if the evolutionary dynamics of internal replicator systems are slow , the internal replicator coexistence is deterministically stable; hence , strengthening the deterministic flow of the internal replicator dynamics would increase the longevity of vesicles . This point seems to be generally relevant in compartmentalized interacting replicator systems ( i . e . the systems where replicases catalyze the replication of templates ) . The crucial difference between this study and our previous study [30] lies in the type of replicator systems considered: the current study considered a compartmentalized interacting replicator system , whereas the previous study considered a compartmentalized non-interacting replicator system ( i . e . the systems where templates self-replicate ) . Since the evolutionary dynamics of the non-interacting replicator system is , in principle , stable under well-mixed conditions [13] , [26] , [27] , the death of vesicles hardly happened unless externally introduced in the previous study . Therein we reported that it was essential to set the death rate of vesicles—which was one of the parameters—sufficiently high for the vesicle-level selection to be effective . In contrast , vesicle death in the current model was not only an internal—or spontaneous—process due to parasites ( i . e . no need to externally introduce vesicle death ) , but also more frequent to those that contain more severe parasites , which reinforces the vesicle-level selection . Interestingly , therefore , parasites , which were the very reason the higher-level selection had to be considered , actually made the vesicle-level selection more effective . We add that we deliberately avoided making a quantitative comparison between the models with respect to the area of the survival region in the parameter space and the maximum tolerable mutation rate for the following reasons . Firstly , the models have qualitatively different kinds of selection pressure because of the functionality of the folded state of parasites . Secondly , the result of a quantitative comparison depends on the parameters ( e . g . one model can have a greater or smaller value of depending on the value of ) . Thirdly , the models do not have a completely identical set of parameters ( most prominently , it is unclear how to scale and ) . Fourthly , there are two different kinds of population size , i . e . that of microscopic entities and that of mesoscopic entities , which can change through evolution ( e . g . Fig . 9 ) . These points make the definition of fairness in quantitative comparison impracticable . Therefore , we concentrated on the qualitative comparison . We should also mention an important simplification made in the current models; i . e . , mutations of replicators were restricted to the perturbation of the two parameters of parasites . Other types of mutation processes can have significant impacts on the eco-evolutionary dynamics of replicator systems . The diversity in the replicase population [30] and/or deleterious mutations [28] can disadvantage parasites by effectively “diluting” the replicase population . Moreover , the explicit consideration of genotype-phenotype-interaction mapping allows a positive feedback in the evolution of these three levels , which stabilizes the whole system [35] . Hence , subjecting a greater degree of freedom to evolution seems to have positive effects on the stability of the replicator systems . In the current study , however , we restricted these processes for a clear-cut elucidation of the effects of different multilevel selection mechanisms . Finally , let us comment on an interesting difference between the modern cell and the protocell conceived in this study ( i . e . the vesicle containing replicators ) . The difference lies in the concept of genotype , which , we commonly assume in evolutionary biology , is a static state of an individual . Such an assumption can be justified for a modern cell because of the small rate of somatic mutation relative to the lifetime of the cell . However , it is clearly invalid for the protocell in the current study because the internal replicator system—of which population composition can be considered as the “genome” of the protocell—greatly changes its state over time comparable to the lifetime of a vesicle ( Fig . 4 ) . Stated differently , one cannot separate from each other the timescale of the eco-evolutionary dynamics of the microscopic entities and the population dynamics of the mesoscopic entities . It would be an interesting future project to investigate how these two timescales can be split apart through evolution ( see also [] ) . There are ongoing efforts to synthesize chemical systems that can undergo self-sustained Darwinian evolution in the laboratory . In particular , Szostak et al . have been making steady progress towards the laboratory synthesis of model protocells[31] , [32] , [49]–[53] . In these studies , the diameter of vesicles ranged from to . Assuming that the vesicle dynamics can keep pace with the replicator dynamics , the number of replicators inside the vesicle should be about 100 to counter the evolutionary instability of the kind of replicator systems investigated in this study ( Fig . 12B ) . Thus , the concentration of polynucleotide inside the vesicle should be for and for , where it is assumed that vesicles are spherical and unilamellar ( for multilamellar vesicles , the greater concentration would be allowed ) . Next , there are experimental techniques to amplify polynucleotide in diffusion limited media , the so-called “molecular colony” or “polony” technology [54] , [55]; and its possible use for chemical Darwinian systems has been suggested [56] . Under the condition in which this technique is normally practiced , it can be calculated that within a molecular colony a volume of contains about 4000 polynucleotide molecules of about 100 bases and that it takes about 10 sec [which we assume is a generation time ( ) of replicators] for diffusion to displace a molecule of a similar length by ( see also Text S1 note 6 ) . The corresponding number of molecules is about 10 for ( 100 for ) in the current surface model . Thus , to obtain a similar situation under the molecular colony technology , either the concentration within a colony must be reduced to 0 . 1 nM or the diffusion constant must be reduced to ( but see Text S1 note 7 ) . Unfortunately , these figures are not applicable to the best currently available RNA-directed RNA polymerase ribozyme ( a candidate of the replicators investigated here ) , because it is at best three order of magnitude slower polymerase than the protein polymerases used for PCR [57] . Moreover , mineral surfaces have been suggested to have various chemical advantages for the origin of life ( e . g . [49] , [58] , [59] , [60] ) . Besides chemical aspects , it appears to us that mineral surfaces also have an advantage in the current context by confining the replicator in two-dimensional space . However , to our knowledge , experimental data necessary for the type of calculation made above seem to be not yet available . Finally , Koonin and Martin recently discussed a system compartmentalized by inorganic boundaries which are static relative to the internal replicator dynamics [61] ( it is reported that related experimental work is in progress [62] ) . The current model can easily be extended to simulate a simplified version of such a system by disabling the growth and division of vesicles and by allowing the small diffusion of replicators across compartment boundaries . Our preliminary investigation showed that a model with static compartments displayed the formation of the large traveling waves typically spanning more than 10 compartments ( where was fixed either to 100 or to 1000 , and inside compartments , and across compartment boundaries ) , which gives rise to the evolutionary dynamics on the level of waves . Thus , interestingly , it is spatial self-organization that plays an important role for the macroscopic stability of the static compartment model , despite the fact that the system is explicitly compartmentalized . Traditionally , multilevel selection has been investigated in the context of altruism-egoism dichotomy . In this context , models are constructed by defining the traits ( or strategies ) of individuals directly with respect to its fitness contributions at different levels of biological organizations either through a priori conception or through inference from observation as such ( e . g . [63] , for review; see also Text S1 note 8 ) . By using these models , the classical theory asserts that a trait that disadvantages an individual can still evolve if it advantages certain higher-level biological organizations . By contrast , the models in the current study were constructed without preconceiving what is costly or beneficial on what level of organization ( see also Text S1 note 9 ) . Therewith , this study—while concordant with the classical theory—gives two novel implications . Firstly , interactions between the dynamics of microscopic and mesoscopic entities can generate novel evolutionary directions ( or strategies ) not conceived in the altruism-egoism dichotomy . Secondly , difference in mesoscopic entities can lead to difference in the long-term evolutionary trend of otherwise identical microscopic entities . Hence , we suggest that it is necessary to go beyond the classical modeling framework in order to explore a possible plethora of novel evolutionary directions—beyond that found here and in [41] , [43]—that can be generated through multilevel evolution . | The origin of life has ever been attracting scientific inquiries . The RNA world hypothesis suggests that , before the evolution of DNA and protein , primordial life was based on RNA-like molecules both for information storage and chemical catalysis . In the simplest form , an RNA world consists of RNA molecules that can catalyze the replication of their own copies . Thus , an interesting question is whether a system of RNA-like replicators can increase its complexity through Darwinian evolution and approach the modern form of life . It is , however , known that simple natural selection acting on individual replicators is insufficient to account for the evolution of complexity due to the evolution of parasite-like templates . Two solutions have been suggested: compartmentalization of replicators by membranes ( i . e . , protocells ) and spatial self-organization of a replicator population . Here , we make a direct comparison of the two suggestions by computer simulations . Our results show that the two suggestions can lead to unanticipated and contrasting consequences in the long-term evolution of replicating molecules . The results also imply a novel advantage in the spatial self-organization for the evolution of complexity in RNA-like replicator systems . | [
"Abstract",
"Introduction",
"Models",
"Results",
"Discussion"
]
| [
"computational",
"biology/evolutionary",
"modeling",
"evolutionary",
"biology",
"chemical",
"biology/directed",
"molecular",
"evolution"
]
| 2009 | Multilevel Selection in Models of Prebiotic Evolution II: A Direct Comparison of Compartmentalization and Spatial Self-Organization |
Growing experimental evidence indicates that , in addition to the physical virion components , the non-structural proteins of hepatitis C virus ( HCV ) are intimately involved in orchestrating morphogenesis . Since it is dispensable for HCV RNA replication , the non-structural viral protein NS2 is suggested to play a central role in HCV particle assembly . However , despite genetic evidences , we have almost no understanding about NS2 protein-protein interactions and their role in the production of infectious particles . Here , we used co-immunoprecipitation and/or fluorescence resonance energy transfer with fluorescence lifetime imaging microscopy analyses to study the interactions between NS2 and the viroporin p7 and the HCV glycoprotein E2 . In addition , we used alanine scanning insertion mutagenesis as well as other mutations in the context of an infectious virus to investigate the functional role of NS2 in HCV assembly . Finally , the subcellular localization of NS2 and several mutants was analyzed by confocal microscopy . Our data demonstrate molecular interactions between NS2 and p7 and E2 . Furthermore , we show that , in the context of an infectious virus , NS2 accumulates over time in endoplasmic reticulum-derived dotted structures and colocalizes with both the envelope glycoproteins and components of the replication complex in close proximity to the HCV core protein and lipid droplets , a location that has been shown to be essential for virus assembly . We show that NS2 transmembrane region is crucial for both E2 interaction and subcellular localization . Moreover , specific mutations in core , envelope proteins , p7 and NS5A reported to abolish viral assembly changed the subcellular localization of NS2 protein . Together , these observations indicate that NS2 protein attracts the envelope proteins at the assembly site and it crosstalks with non-structural proteins for virus assembly .
The hepatitis C virus ( HCV ) has a high propensity to establish a persistent infection in the human liver . Approximately 170 million people suffer from chronic hepatitis C and are at risk to develop cirrhosis and hepatocellular carcinoma [1] . Current antiviral therapy is based on the use of polyethylene glycol conjugated interferon alpha in combination with ribavirin . However , this treatment is expensive , relatively toxic and effective in only approximately half of the treated patients [2] . A better understanding of the HCV life cycle is therefore essential for the development of more efficacious and better tolerated anti-HCV treatments . HCV is an enveloped virus that belongs to the Hepacivirus genus in the Flaviviridae family [3] . HCV has a positive strand RNA genome encoding a single polyprotein that is cleaved by cellular and viral proteases into 10 different proteins: core , E1 , E2 , p7 , NS2 , NS3 , NS4A , NS4B , NS5A and NS5B [3] . The non-structural proteins NS3 to NS5B are involved in the replication of the viral genome , whereas the structural proteins ( core , E1 and E2 ) are the components of the viral particle ( reviewed in [4] ) . The remaining proteins , p7 and NS2 , are dispensable for RNA replication and there is no evidence that they are part of the viral particle [5] , [6] . For reasons still unknown , HCV clinical isolates do not propagate in cell culture . However , with the development of a cell culture system that enables a relatively efficient amplification of HCV ( HCVcc ) [7] , [8] , [9] , all the steps of the HCV life cycle can be investigated . Due to the accumulation of HCV core protein around lipid droplets ( LDs ) ( reviewed in [10] ) , a role of these lipid bodies in HCV assembly has been suspected for a long time . Moreover , it has recently been shown that viral non-structural proteins like NS5A and NS3 and double stranded viral RNA are also present around LDs [11] , [12] . The association between core and LDs seems to play a role in the recruitment of the other viral proteins and for virus production [12] , [13] . Furthermore , NS5A plays a double role in both replication and assembly processes as a potential switch between these two steps [14] , [15] , [16] , [17] . Since it is dispensable for HCV RNA replication and it does not seem to be incorporated into viral particles [6] , NS2 has been suspected to be involved in the assembly process of HCV particle . Recently , experimental evidence supporting this hypothesis has been obtained [18] , [19] , [20] . Although the role of NS2 in the assembly process remains elusive , some data suggest that NS2 might interact with viral partners involved in virion morphogenesis . Indeed , construction of chimeric viruses between different genotypes identified the C-terminus of the first transmembrane segment of NS2 as the optimum crossover point [21] . Thus , a genetic interaction was implied between the N-terminus of NS2 and the upstream structural proteins . In the context of a chimeric virus containing genotype 1a and 2a sequences , adaptive mutations in E1 , p7 , NS2 and NS3 were identified , also suggesting genetic interactions between these proteins [22] . Moreover , a detailed rescue mutant analysis recently showed genetic interactions between NS2 , E1E2 and NS3-4A [23] . However , despite these genetic analyses we have almost no understanding about NS2 interactions with other viral proteins and the role of these interactions in the production of infectious particles . Here , we report molecular interactions between NS2 and p7 and E2 proteins . Using a functional HCVcc virus with a reporter epitope at the N-terminus of NS2 , we found that NS2 accumulates in dotted structures derived from the endoplasmic reticulum ( ER ) and colocalizes with E1 , E2 , NS3 and NS5A in close proximity to the core protein and LDs . Mutations and deletions in the p7-NS2 region affecting the subcellular localization of NS2 and its physical protein-protein interactions abolished viral assembly . Moreover , mutations in other viral proteins reported to inhibit the assembly process induced consistent changes in NS2 subcellular localization . Together , these data suggest that p7 , NS2 and E2 form a functional unit which drives the proteins in the proximity of the LDs where NS2 crosstalks with other viral proteins during the virion assembly process .
NS2 is a polytopic transmembrane protein containing 3 putative transmembrane segments [19] ( Figure 1 ) . The p7 polypeptide and E1E2 heterodimer , which are putative partners of NS2 , are also membrane proteins that contain transmembrane segments [24] , [25] . Due to their respective topologies ( Figure 1A ) , it is expected that interactions between these three proteins would involve helix-helix contacts in their transmembrane segments . Furthermore , helix-helix interactions between transmembrane segments of NS2 are also likely to take place . To analyze the role of the transmembrane domain of NS2 in potential protein-protein interactions , we used a previously reported deletion mutant of NS2 ( ΔTM12 ) which lacks the first two transmembrane segments [19] . For a more refined analysis , we also used alanine scanning insertion mutagenesis , a technique which has been shown to disrupt helix-helix interactions in a membrane environment [26] , [27] , [28] , [29] . As illustrated in Figure 1C for the first transmembrane segment of NS2 , this approach is based on the fact that insertion of a single amino acid into a transmembrane helix displaces the residues on the N-terminal side of the insertion by 110° relative to those on the C-terminal side of the insertion . The subsequent perturbation of the residue side-chain distribution could disrupt a potential helix-helix packing interface involving residues on both sides of the insertion . Such mutations are therefore expected to disrupt helix-helix interactions between the transmembrane segments of NS2 and/or between NS2 and other putative partners like p7 or E1E2 . Based on the NMR structure of the transmembrane segments of NS2 [19] ( Jirasko et al . , 16th International Symposium on HCV and Related Viruses , Nice , October 3–7 , 2009 ) , we designed insertion mutations in the transmembrane domain of NS2 . Three different mutants were designed by inserting alanine residues close to the middle of the transmembrane segments of NS2 ( Figure 1 , B–C ) . The positions for alanine insertions were carefully chosen within the putative helical segments to preserve the overall fold of the corresponding helices . According to previous reports [26] , [27] , [28] , [29] , alanine insertions near the center of putative transmembrane helices were expected to be the most efficient to disrupt inter-helices potential interactions . Firstly , we wanted to evaluate the impact of our mutations on the viral life cycle . As negative controls for virus assembly , we used assembly-deficient viruses JFH-ΔE1E2-HA and JFH-Δp7-HA , containing a deletion in the regions encoding HCV envelope glycoproteins and the p7 polypeptide , respectively ( Figure S1 ) . As template for all our constructs , we used a full-length JFH1 plasmid containing adaptive mutations [30] in which the N-terminal sequence of E1 has been modified to reconstruct the A4 monoclonal antibody ( Mab ) epitope of the H77 isolate in order to facilitate the immunodetection of this envelope protein [31] . Moreover , we introduced an HA epitope in the N-terminus of NS2 to be used for protein detection ( JFH-HA ) . Huh-7 cells were electroporated with different mutated viral genomes and the production of infectious virus was assessed at 72h by supernatant titration ( Figure 2A ) . The insertion of the HA epitope did not affect the virus production as compared to the wild-type virus ( JFH ) . As previously reported , deletion of the first two transmembrane segments prevented the production of infectious particles similar to the negative controls [19] ( Figure 2A and S1 ) . Furthermore , alanine insertions also drastically affected the viral production . While mutants A16 and A41 presented residual infectivity , A82 was not able to produce infectious particles ( Figure 2A ) . To identify the stage where the virus production was affected , we started with the evaluation of the replication capacity of our viruses . To this aim , we used Renilla luciferase reporter viruses ( Figure S1A ) . As a negative control for replication , we used the replication-deficient construct JFHGND-Luc which contains a mutation in NS5B that prevents viral genome replication [7] . As the RNA input after electroporation is potentially variable , we evaluated the capacity of our viruses to replicate by determining the ratio between the luciferase activity at 72h and 4h post-electroporation when only the luciferase activity of the input RNA is present as previously shown [32] . As shown in Figure S1B , all the mutant viruses presented a similar replication capacity , which was comparable to the control viruses , whereas replication was abolished in the JFHGND-Luc mutant ( Figure S1B-upper panel ) . Since the replication was not affected , the effect of alanine insertion on NS2 protein stability was further investigated . As shown in Figure S1C , A82 insertion was detrimental for the protein integrity . For A16 mutant , the level of expression of NS2 was slightly lower as compared to the JFH-HA control . However , the level of expression of HCV proteins was higher for the JFH-HA control in this particular experiment . We therefore measured the NS2/E2 ratios which were similar for A16 , A41 and JFH-HA , indicating that A16 and A41 insertions did not affect NS2 stability . Therefore , we focused our analysis on A16 and A41 mutants . To test whether the lack of infectivity could be due to a defect in virus secretion or the release of non-infectious particles , we determined the level of core protein in the supernatants . The quantity of core protein in supernatants decreased drastically as compared to wild-type , paralleling the decrease in infectivity ( Figure 2B ) . It has to be noted that for the mutants showing a residual infectivity ( A16 , A41 and RR/QQ ) , the release of core was at the same level as the dead mutants ( ΔTM12 , A82 and Δp7 ) . This likely reflects a difference of sensitivity between the two assays . Then , we measured the viral RNA in the supernatants by qRT-PCR as previously described [17] . For all the alanine insertions , the release of viral RNA was close to the background level observed for the assembly-deficient control viruses ( Figure S1B – lower panel ) , indicating a defect in particle secretion . These data suggest that either the process of assembly is affected at an early stage or the secretion of assembled infectious particles is impaired . To answer this question , we compared the intra and extracellular level of the core protein ( Figure 2B ) . The ratio between intra and extracellular core was similar to JFH-Δp7-HA which was shown to be defective in early assembly steps . We also measured the intracellular infectivity of the alanine mutants in the Rluc reporter viruses context as previously reported [20] . As shown in Figure S1B ( middle panel ) , there was no accumulation of intracellular infectivity for any of the mutants . These results suggest that mutations in the transmembrane domain of NS2 prevent the assembly process rather than the secretion of particles . Considering the rationale of our mutagenesis , we investigated protein-protein interactions in E1E2-p7-NS2 region . Indeed , due to their position within the HCV polyprotein , it is reasonable to think that these proteins might potentially interact . Genetic and co-immunoprecipitation data suggest that NS2 interacts with E2 [23] , [32] , [33] , [34] . Interaction assays between NS2 and E2 were performed in the context of the JFH-HA virus , which allows to immunoprecipitate NS2 with the HA tag ( Figure 2C , JFH-HA ) . As shown in Figure 2C ( JFH-HA ) , E2 and NS2 proteins migrated at the expected molecular mass , indicating that the polyprotein was correctly processed in these viruses . It has to be noted that in the absence of p7 , E2 migrated slightly faster . This is likely due to the absence of E2-p7 , which has a slightly slower migration profile in the band corresponding to E2 . The E2-NS2 interaction was then tested by co-immunoprecipitation with an anti-HA antibody followed by the detection of E2 by Western blotting . As shown in Figure 2C ( JFH-HA ) , E2 co-precipitated with NS2 , confirming that these proteins interact together in the context of the virus . In contrast , deletion of the first two transmembrane segments prevented the interaction between NS2 and E2 ( Figure 2C , ΔTM12 ) , whereas alanine insertions in the transmembrane region had different effects . While A16 ( within TM1 ) did not affect the E2-NS2 interaction , A41 ( within TM2 ) induced a consistent decrease in the amount of E2 co-precipitated by NS2 ( Figure 2C , A16 and A41 ) . We further investigated the effect of p7 on the E2-NS2 interaction . To this aim , we used two mutants of p7 , a deletion mutant ( JFH-Δp7-HA ) and a mutant having the two arginine residues in the cytosolic loop of p7 replaced by glutamine residues ( JFH-RR/QQ-HA ) ( Figure 2C ) . This RR/QQ mutation is believed to abolish the ion channel activity of p7 , and it induces a drastic decrease in virus production as previously reported [32] , [35] and confirmed by us as shown in Figure 2A . As shown in Figure 2C ( JFH-Δp7-HA and JFH-RR/QQ-HA ) , E2 and NS2 proteins migrated at the expected molecular mass , indicating that the polyprotein was correctly processed in these viruses . However , the two mutations had no significant effect on E2-NS2 interaction ( Figure 2C JFH-Δp7-HA and JFH-RR/QQ-HA ) , suggesting that p7 does not modulate E2-NS2 interaction . To obtain more insight into the assembly defects of our viral mutants , we decided to determine their potential effect on the subcellular localization of NS2 . To this aim , we first characterized the subcellular localization of the wild-type NS2 in the context of an infectious virus , which has never been reported before . NS2 presented an ER-like reticulated pattern , and in some cells , accumulated in dotted structures ( Figure 3A and data not shown ) . The number of these NS2-positive dot-like structures increased over time , suggesting a transition from an initial reticulate pattern to these dotted structures . At 72 hours post-electroporation , 34±15% of infected cells from 8 different electroporations displayed NS2 dots . The size of these structures was 0 . 84±0 . 38 µm ( mean±SD , n = 402 ) . To further characterize these NS2 structures , we performed co-localization analyses with different cellular markers . As expected , the reticulate and perinuclear pattern of NS2 overlapped with calreticulin staining as well as other ER markers like calnexin and PDI ( data not shown ) . Interestingly , NS2 dots also colocalized with ER markers ( Figure 3B ) , but they did not colocalize with other organelle markers of the secretory pathway ( data not shown ) . Importantly , NS2 dots were also found in close proximity of LDs ( Figure 3B ) , suggesting that they might play a role in HCV assembly . To better understand the potential role of NS2 dots in virus assembly , we analyzed the subcellular localization of NS2 in relationship with the other viral proteins . NS2 dots overlapped with HCV envelope glycoproteins E2 ( Figure 3B ) and E1 ( Figure S2 ) . NS5A and NS3 were also shown to colocalize with NS2 in its dotted pattern ( Figures 3B and S2 ) . Importantly , the structures containing both NS2 and NS5A were observed in close proximity to LDs ( Figure 4A ) . This is illustrated by the presence of magenta dots ( red NS2 and blue NS5A ) in the proximity of LD ( green ) . Finally , as observed with the LDs , NS2 dots were also found in close proximity to core protein ( Figure S2 ) , and as expected this association was observed in close proximity to LDs ( Figure 4B ) . As shown in Figure 4B , core ( blue ) is tightly associated to LD ( green ) as the LD becomes cyan due to colocalization , NS2 ( red ) localizes in regions juxtaposed to the cyan LD ( Figure 4B ) . Relying on the spatial proximity of NS2 dots , core , E1E2 , NS3 , NS5A proteins and LDs , we speculated that NS2 present in these structures is involved in the assembly process . Thus , we established some criteria of functionality for NS2 positive structures . They have to localize in the proximity of LDs and core protein and more importantly to colocalize with NS5A protein , which we used as a criteria for quantification purposes . It has to be pointed out that a low number of cells contained NS2-positive structures with a different pattern of subcellular localization ( Figure S3 ) . This different pattern was indeed observed in approximately 1 to 3% of the cells at 72h post-infection or post-electroporation . These structures colocalized less with ER markers and they overlapped with ERGIC-53 , a marker of the ER-to-Golgi intermediate compartment [36] ( Figure S3 ) . Furthermore , these NS2 dots were detected in close proximity to the ER exit sites , which were identified by markers of the COP II coatomer , Sec31 and p125 [37] , [38] ( Figure S3 and data not shown ) . However , cells containing these NS2 positive structures showed dramatic alterations of the secretory compartments as observed by immunofluorescence analysis of ER-to-Golgi intermediate compartment and Golgi morphology using ERGIC-53 and GM130 as markers ( Figure S3 ) . It is worth noting that NS2 did not colocalize with NS3 or NS5A and it was not found in the proximity of core and LDs in these cells ( Figure S3 ) . Due to these alterations , it is unlikely that these cells are involved in the production of infectious virus . We asked further the relevance of NS2 dots for the production of infectious particles . After electroporation , we determined the titer of virus production at different time points ( Figure 4C , upper panel ) . In parallel , we counted the number of cells which presented NS2/NS5A positive dots for reasons detailed above ( Figure 4C , lower panel ) . The kinetics of virion production and the percentage of cells presenting NS2/NS5A positive dots showed a high correlation with a correlation coefficient of 0 . 9 . To exclude the possibility that the NS2 phenotype depends on the cell culture adaptive mutations , we performed a similar experiment with a virus that does not contain the mutations . As for the adaptive mutant , virus production in the absence of mutation paralleled the NS2/NS5A dots formation with a significant correlation coefficient ( data not shown ) . These data reinforce the idea that NS2/NS5A positive dots are involved in the virus production process . To further investigate the NS2 localization , we performed immuno-electron microscopy with an anti-HA antibody on Huh7 cells which were electroporated with JFH-HA RNA and prepared by cryosubstitution . As shown in Figure 4D , we could identify clusters of gold particles in the proximity of LDs . Moreover , two of the clusters are lying on preserved ER bilayers . For one of the clusters of gold particles , a connection between the ER bilayer and the LDs could be observed ( Figure 4D ) . Thus , the immuno-electron microscopy confirms the juxtaposition between NS2 dots and LDs in a 0 . 2µm range , which is consistent with the observations in confocal microscopy . It has to be noted that gold particles were detected on both sides of the membrane even if the HA epitope is supposed to be located in the ER lumen . This is compatible with the length of two antibodies ( primary+secondary ) since the gold particles were never further away than 30 nm from the membrane . However , we cannot exclude a double topology for NS2 as recently suggested [39] . The next obvious step was to determine the subcellular localization of NS2 for the different mutants defective in assembly . In the case of ΔTM12 , NS2 localized in confined structures which did not colocalize with NS5A and they were not associated with the core protein or LD ( Figure 5 , panels A and C and data not shown ) . Moreover , there was no colocalization between the truncated NS2 and the E2 glycoprotein ( Figure 5B ) , which correlates with the lack of interaction between the two proteins ( Figure 2C ) . Then , we analyzed the subcellular localization of NS2 for the alanine insertion mutants . While A16 mutant presented NS2 dotted structures as wild-type , A41 mutation induced a drastic decrease in the percentage of cells with NS2 dots ( Figure 5 , panels A and C ) . These data suggest that NS2 transmembrane region is an important localization determinant . A peculiarity of the Flaviviridae family is the involvement of both structural and non-structural proteins in the assembly process ( reviewed in [40] ) . Thus , we wanted to investigate the NS2 subcellular localization in the context of assembly deficient viruses having mutations in different viral proteins . Recruitment of core protein to LDs was reported to be essential for a productive assembly process [12] , [13] . The proline residues 138 and 143 in domain D2 of the core protein are crucial for virus production and core recruitment to LDs [13] , [41] . Furthermore , the mutation of these proline residues has been previously shown to prevent the core induced recruitment of NS5A to the LDs [12] . Therefore , we introduced these mutations in the context of JFH-HA virus ( Figure 6 , JFH-HA-PP ) . As previously reported [13] , the mutation prevented the production of infectious virions ( Figure 9A ) , and the core protein was not redistributed to LDs , which in turn remained spread in the cytoplasm rather than the perinuclear localization induced by a functional core protein ( Figure S4 ) . As shown in Figure 6 , in the context of this mutation , NS2 protein maintained the capacity to accumulate in dotted structures that colocalized with NS5A . In contrast to the wild-type , NS2 dotted structures were not found in the vicinity of LDs in the context of the PP mutation , suggesting that NS2 does not have the signals to localize by itself around the LDs ( Figure S4 ) . Importantly , in this context , the number of cells presenting NS2 dotted structures increased tremendously in comparison to the wild-type ( Figure 6B ) . These observations suggest that the PP mutation induces a block in the assembly process , which favors the accumulation of NS2 protein in the dotted structures . As structural proteins , the envelope glycoproteins are involved in the assembly process [42] . The envelope proteins play a crucial role in the assembly of enveloped viruses , which can be due for some viruses to the capacity of the envelope proteins to establish lateral interactions and to generate a pushing force necessary for the budding process [43] . A deletion in the envelope region would therefore block the assembly process as shown by an in-frame deletion of 351 amino acids in the envelope proteins region [7] . In the context of our JFH-HA virus , this deletion mutant is also fully replicative , and it does not produce infectious particles ( data not shown ) . To further understand the interplay between NS2 and HCV envelope glycoproteins , we analyzed the subcellular localization of NS2 protein in the context of the E1E2 deletion . In this context , NS2 localized in NS5A positive structures juxtaposed to the LDs and core protein ( Figure 6 and S4 ) . Interestingly , as for the JFH-HA-PP virus , the number of cells presenting NS2 dotted structures also increased in the case of JFH-ΔE1E2-HA , correlating also with a block in the assembly process favoring an accumulation of NS2 dotted structures ( Figure 6B ) . The deletion introduced in the envelope region is predicted to generate a chimeric protein comprising the N-terminus of E1 and the C-terminus of E2 protein . Since we introduced the A4 epitope in the N-terminus of E1 , we were able to detect the subcellular localization of this small chimeric protein . It is worth mentioning that we also detected this truncated chimeric protein in NS2/NS5A dotted structures , suggesting that the E2 transmembrane domain is sufficient for the recruitment of HCV envelope proteins to NS2 dotted structures ( Figure S4 ) . The p7 protein has been shown to be crucial for the assembly process [32] , [35] . Moreover , the ion channel activity of p7 correlates with the virus assembly process since mutations predicted to abolish the ion channel activity have a strong inhibitory effect on the virus production [32] , [35] . We therefore also analyzed the subcellular localization of NS2 in the context of p7 mutants corresponding to a complete deletion ( JFH-Δp7-HA ) or an amino acid substitution ( JFH-RR/QQ-HA ) previously reported to affect the assembly and release of the virus [32] , [35] . The two constructs behaved as expected . While JFH-Δp7-HA produced no infectious particles , JFH-RR/QQ-HA presented a 2 log10 decrease in virus titers at 72h post-electroporation ( Figure 2A ) . Importantly , the two mutants induced a drastic decrease in the number of cells presenting NS2/NS5A dotted structures ( Figure 6B ) . Together , these data indicate that NS2 needs a functional p7 polypeptide to colocalize with NS5A in dotted structures . Other partners than p7 are likely necessary for NS2 accumulation in dotted structures . Indeed , as shown in Figure 5 , A41 and ΔTM12 mutants , which fail to interact with E2 , also present a drastically reduced number of NS2 dotted structures . This suggests that NS2-E2 interaction might be crucial for NS2 subcellular localization . Thus , one additional determinant could be represented by the transmembrane domain of E2 , which is most likely the interacting region with NS2 due to topological constraints . In order to check this hypothesis , we used both JFH-HA and JFH-ΔE1E2-HA constructs in which we replaced the transmembrane region of E2 with the autoprotease 2A from foot and mouth disease virus ( FMDV ) ( JFH-ΔTME2-HA and JFH-ΔE1E2TME2-HA ) ( Figure S1 ) . We first verified that a proper processing of the polyprotein mediated by FMDV 2A protease has occurred in these constructs . To check our constructs , we analyzed the molecular mass of E2 following deglycosylation with EndoH or PNGase endoglycosidases . As shown in Figure 6C , E2 from JFH-HA and JFH-ΔTME2-HA presented a similar molecular mass after deglycosylation with EndoH , suggesting that the FMDV 2A protease is functional since FMDV 2A and the transmembrane domain of E2 have similar sizes . Indeed , if FMDV 2A protease had not been functional , we would have observed a difference of 7kD corresponding to the molecular mass of unprocessed p7 ( Figure 6C and data not shown ) . Interestingly , in contrast to the wild-type envelope protein , the truncated E2 did not interact with NS2 ( Figure 6D ) . Furthermore , in contrast to what was observed for JFH-HA and JFH-ΔE1E2-HA ( Figure 6A ) , NS2 protein of the ΔTME2 mutant presented an ER like pattern and the formation of NS2 dotted structures was prevented ( Figure 6A and B ) . Similar data were also obtained with JFH-ΔE1E2TME2-HA construct ( Figure 6B ) . Thus , it seems that p7-NS2 and the transmembrane domain of E2 form a functional unit that targets these proteins to NS5A positive structures . The above data suggest a possible interaction between p7 and NS2 . We therefore explored this putative interaction in a biochemical assay , by analyzing p7-NS2 association in a co-immunoprecipitation assay . Due to the difficulties in analyzing p7-NS2 interactions in the context of an infectious virus , we analyzed these interactions by co-transfecting cells with plasmids expressing these two proteins only . In this approach , the p7 polypeptide and NS2 were tagged with a Flag or a HA epitope , respectively ( Figure 7A , p7-Flag and HA-NS2 ) . The p7-NS2 interaction was tested after co-expression of the tagged proteins in 293T cells . Co-immunoprecipitation experiments were performed with an anti-Flag antibody linked to agarose beads . The immunoprecipitates were separated by SDS-PAGE and probed with anti-HA antibodies by Western blotting . As shown in Figure 7B , NS2 of different genotypes coprecipitated with p7 . Since only p7 of genotype 1a was used in these experiments , it suggests that p7 interacts with NS2 in a genotype independent manner . However , we cannot exclude that the system is not sensitive enough to discriminate between slight changes in affinities . Further , we constructed two chimeric proteins , NS2 tagged with a green fluorescent protein ( NS2-GFP ) and NS2-GTM ( Figure 7A ) . In NS2-GFP , the cytosolic domain of NS2 was replaced by GFP protein , whereas for NS2-GTM , we replaced the transmembrane domain of NS2 by the transmembrane domain of glycoprotein G of VSV . As shown in Figure 7B , the NS2-GFP could be precipitated by p7-FLAG , while NS2-GTM could not . This clearly shows that the transmembrane region is the main determinant of p7-NS2 interaction . To confirm the p7-NS2 interaction with another approach , we used the FRET-FLIM technique . FRET-FLIM requires the presence of two fluorophores ( a donor and an acceptor ) fused in frame to the studied proteins . If the two proteins interact , an energy transfer occurs between the two fluorophores and the fluorescence life time of the donor ( a parameter of the energy emitted by the donor ) will decrease . To measure p7-NS2 interactions by FRET-FLIM , Cerulean fluorescent protein ( CFP ) and Venus yellow fluorescent protein ( YFP ) were fused to the N-terminus of p7 and NS2 , respectively ( Figure 8A ) . As previously reported , CFP-p7 and YFP-NS2 showed a reticulate perinuclear distribution ( Figure 8B ) , which is characteristic of ER proteins [34] , [44] . Western blotting analyses indicated that CFP-p7 and YFP-NS2 migrate at the expected molecular mass with some degradation products of lower molecular weight ( Figure 8C ) . The energy transfer in FRET-FLIM assay needs the integrity of the fluorophores and the correct positioning of the interacting partners . The degradation products could fall into two categories: either soluble fluorophores or membrane bound truncated chimeras . In either case the energy transfer is unlikely to occur with the cleavage products . Thus , the presence of degradation byproducts is unlikely to influence the accuracy of the FRET-FLIM acquisitions . After biphoton laser excitation and data analysis , fluorescence life time maps were built . Interestingly , the regions showing interactions were located in distinct spots throughout the cells as illustrated in a fluorescence life time color-coded map ( Figure 8D ) . A summary of FRET-FLIM analysis is presented in Table 1 . The mean life time of fluorescence decreased from 2 . 69±0 . 12 ns ( n = 10 ) in cells transfected with the donor only ( CFP-p7 ) to 2 . 34±0 . 09 ns ( n = 10 ) for double transfections ( CFP-p7+YFP-NS2 ) . The variation of the mean donor life time is characteristic for energy transfer between two fluorescent proteins in FRET-FLIM analyses as previously observed for other protein-protein interactions [45] , [46] . As a positive control , we used the transmembrane domains of HCV glycoproteins E1 and E2 which are known to interact and to have the same subcellular localization as p7 and NS2 [25] . The positive control couple presented a comparable decrease in the mean lifetime of the donor to CFP-p7/YFP-NS2 couple . Indeed , the mean life time decreased from 2 . 56±0 . 03 ns ( n = 14 ) in cells transfected with the donor only ( CFP-E2 ) to 2 . 35±0 . 08 ns ( n = 14 ) for double transfections ( CFP-E2+YFP-E1 ) . As a negative control , we used CFP fused to the transmembrane domain of yellow fever virus E protein ( CFP-EYF ) , a donor protein with the same topology and localization as p7 [28] , [47] . As shown in Table 1 , the mean life time of the donor in monotransfection did not change in double transfections confirming the lack of interaction between CFP-EYF and YFP-NS2 ( n = 11 ) . The biphoton pictures for the positive and the negative control are shown in Figure S5 . Thus , these data strongly suggest that p7 and NS2 proteins interact intracellularly . Among the non-structural proteins , NS5A is the most characterized in terms of its role in the assembly process . NS5A is recruited through direct interaction by the core protein around LDs where its domain III is involved in the assembly process potentially by its phosphorylation [14] , [16] , [17] . By deletion mutagenesis , Tellinghuisen et al . identified a cluster of serine residues at positions 452 , 454 and 457 , which are crucial for virus production [17] . Furthermore , by alanine scanning mutagenesis , Masaki et al . reported that the same serine cluster is involved in the direct interaction between NS5A and core protein [16] . While Tellinghuisen et al . reported that serine 457 alone is essential for virus production , Masaki et al . showed that only double mutants had a significant impact on virus production [16] , [17] . The apparent contradiction might be explained by the different viruses and time points for virus production assessment . While Tellinghuisen et al . used a chimeric virus consisting of the structural proteins of J6 strain up to NS2 protein , Masaki et al . used the wild type JFH strain [16] , [17] . Thus , we constructed the two mutants in the context of our JFH-HA virus – JFH-S/A-HA and JFH-3BS/A-HA , respectively . We analyzed the phenotype of the mutants as well as the polyprotein processing . The results fitted the literature with JFH-S/A-HA virus infectivity moderately reduced at 72h and JFH-3BS/A-HA profoundly impaired in infectious virus production , while the replication and protein integrity were unaltered ( Figure 9A , B , C ) . As reported , we showed that JFH-S/A-HA and JFH-3BS/A-HA present less hyperphosphorylated NS5A ( Figure 9C ) . Surprisingly , for both mutants , NS2 localized mainly in an ER-like pattern and the number of cells with NS2/NS5A dots decreased dramatically ( Figure 9D , E ) . The serine 457 may be replaced by an aspartate residue , which mimics a phosphoserine [17] . We therefore introduced the same mutation ( JFH-S/D-HA ) and as reported the virus titers were restored at wild-type levels ( Figure 9A ) . Interestingly , this mutant partially recovered the NS2 subcellular localization both qualitatively and quantitatively ( Figure 9D , E ) . Together , these results suggest that NS5A phosphorylation might stabilize the NS2 dotted structures in the assembly process .
Our understanding of the HCV morphogenesis process is still in its infancy . Different viral components were identified as players in the morphogenesis process . As expected , the structural proteins are essential in the virus makeup [7] . The scenario gets more complicated with the involvement of non-structural proteins in the assembly process . The p7 polypeptide , NS2 , NS3 , NS4B , NS5A were reported to be involved in viral assembly [14] , [17] , [20] , [22] , [32] , [48] . However , the mechanism of the complex interplay between the structural and non-structural proteins towards the virion production is not understood . In this paper , we provide evidence for molecular interactions between NS2 , p7 and E2 , respectively . Furthermore , we show that NS2 accumulates over time in ER-derived dotted structures and colocalizes with the envelope glycoproteins and components of the replication complex in close proximity to the core protein and LDs . Characterized assembly deficient mutants in both structural and non-structural proteins present qualitative and quantitative modifications in NS2 subcellular localization . Indeed , specific mutations within NS2 , p7 or E2 modify the subcellular localization of NS2 and impair virus production . Mutations in core , envelope proteins or NS5A affect the NS2 subcellular localization along with the virus titers . Altogether , these observations indicate that NS2 protein crosstalks with both structural and non-structural proteins during virus assembly . Our data demonstrate a physical interaction between NS2 and p7 . This interaction correlates with the previously reported genetic interactions present in the C-NS2 region [21] , [22] . The HCV p7 polypeptide is a viroporin involved in viral assembly [20] , [32] , [49] . Viroporins represent a class of viral proteins that are involved in the viral morphogenesis process in different and largely unknown manners . Alphavirus 6K interacts with E1 and p62 envelope glycoproteins and is involved in optimal assembly and release of the virion by an unknown mechanism [50] , [51] . The E protein of coronaviruses interacts with the M protein and is crucial for the assembly of virus-like particles and virions [52] , [53] , [54] . To our knowledge , the HCV p7 polypeptide is the first viroporin which interacts with a non-structural protein ( NS2 ) and this might be a peculiarity of the members of the Flaviviridae family where the non-structural proteins are involved in particle assembly [40] . Topologically , the transmembrane domain of NS2 ( NS2TM ) would be the main region available for interactions with the upstream transmembrane proteins p7 and E1E2 . Based on this assumption , we wanted to characterize NS2 interactions with p7 and E2 by deletion and chimeric mutants . Both NS2-E2 interaction and NS2-p7 interaction were mapped in the transmembrane region of NS2 as expected . However , since the lack of p7 does not affect the E2-NS2 interaction , we could imagine that NS2 uses separate domains to interact with p7 and the envelope proteins . Interestingly , when we deleted the transmembrane region of E2 , the NS2-E2 interaction was impaired and the NS2 subcellular localization changed . Altogether , these observations suggest that p7 , the transmembrane domain of NS2 and the transmembrane domain of E2 contain signals which act synergistically to direct the NS2 protein towards the NS5A positive membranes in the LD proximity . The drastic effects of alanine insertion mutagenesis on HCV infectivity reflect more intrinsic effects on NS2 function in virus assembly . As shown for lactose permease , single alanine residue insertions into transmembrane helices of a polytopic membrane protein can be highly disruptive to protein structure and function due to their effects on intramolecular helix-helix interactions [26] . Furthermore , within the same protein , different transmembrane helices can have differential sensitivities to single residue insertions [26] . In the case of NS2 , our data indicate that an insertion in the third putative transmembrane helix strongly reduces the stability of the protein , suggesting a drastic alteration of NS2 structure by this mutation . The decrease in infectivity for the mutations in the first two transmembrane helices also indicates a drastic alteration in NS2 function that can be linked to local alteration of its structure , as suggested by the change in subcellular localization of NS2 mutant A41 . Finally , the drastic decrease in infectivity of mutant A16 in spite of NS2 localization in dotted structures indicates that the subcellular localization of NS2 in these structures is not sufficient by itself for infectivity . Rather , it likely needs to play additional function ( s ) at the site of virus assembly and such function ( s ) would be disrupted by the A16 mutation . One explanation could be that a weak E2-NS2 interaction as seen for A41 is not able to direct the p7-NS2 unit to the LDs and a potential strong interaction as for A16 does not allow the p7-NS2 unit to release the envelope proteins heterodimer to the assembly site and the subsequent recycling/dissociation of the unit . Over time , a substantial part of NS2 accumulates in dotted structures localized in the ER in close proximity to the core protein which is associated to LDs . Since the LD/ER interface is considered as the potential particle assembly site [12] , NS2 localization close to the LDs is expected to correlate with its function in a late step of the viral life cycle . Interestingly , NS2 accumulation in dotted structures parallels the colocalization of NS2 with NS5A , NS3 proteins and most likely the replication complex . Moreover , the NS2 dotted structures are juxtaposed to the core protein and colocalize with the envelope proteins E1 and E2 . Thus the NS2 dots contain all the assembly players and are located in the microenvironment of the LDs , the proposed virus assembly site . It is worth mentioning that the formation of NS2 dotted structures is not due to a non-specific effect of the viral genome replication since some of our mutants showing the same replication rate had very different subcellular localizations of NS2 ( e . g . A16 vs . A41 ) . It seems rather that the formation of NS2 related dots represents a transition rather than an end state in a productive assembly process . Indeed , mutations in core ( JFH-HA-PP ) or in the envelope proteins ( JFH-ΔE1E2-HA ) induced an obvious increase in the number of cells presenting NS2 dots . This could mean that the p7-NS2-E1E2 complexes pre-exist in the NS5A positive subcompartment . In addition , the core-mediated redistribution of LDs could induce the recruitment of the assembly components to LDs followed by the envelope protein incorporation into the virion during the budding process . Finally , after budding , NS2 might relocate to another subcellular compartment or be degraded . This scenario is supported by the fact that NS2 dots accumulate around the LDs and core protein when the envelope proteins are deleted , which is expected to inhibit the lateral interactions between the envelope proteins [42] , preventing the budding process to occur . For the moment , the connection between NS2 and the replication complex ( RC ) is just inferred from genetic data and colocalization in immunofluorescence experiments [23] , [34] ( this report ) . We show here that mutations in NS5A , which are reported to affect the phosphorylation status of the protein , abolish the accumulation of NS2 in dotted structures . Moreover , we could partially restore the phenotype by an aspartate mutant , which mimics phosphoserine . Thus , NS2 colocalization with NS5A is favored by the phosphorylation state of the latter . It is possible that NS5A charged state enhances a potential NS2-RC interaction , which translates in NS2 dot formation . Removing the charges would shift the interaction equilibrium and would affect the virus production kinetics in different extents depending on the number of charges and virus strain . Hence , if serine 457 is replaced , the virus titers are moderately reduced at 72h . In contrast , removing serine residues 452 , 454 , 457 determines a profound defect in infectious virus production . However , there is no NS2 accumulation for either of the two mutants . Thus , the NS2 dots may represent transition states in the assembly process . Indeed , some mutations may prevent the arrival of dots components to NS5A structures ( JFH-ΔTM12-HA , JFH-A41-HA , JFH-Δp7-HA and JFH-ΔTME2-HA ) . Alternatively , other mutations may block the assembly and stabilize them ( JFH-HA-PP , JFH-ΔE1E2-HA ) . If we combine a mutation from the former category ( JFH-ΔTME2-HA ) with one from the latter ( JFH-ΔE1E2-HA ) in the mutant JFH-ΔE1E2TME2-HA , we prevent the NS2 dots formation . This suggests that the formation of NS2 complexes and their arrival to the NS5A structures precede the accumulation of NS2 around the LDs during the assembly process . Furthermore , changes in the phosphorylation state of NS5A could regulate the stability of NS2 dots and virion production efficacy . In our current view , the assembly process would involve several steps . Upon viral genome translation and polyprotein processing , formation of different complexes occurs: the E1E2 native heterodimer , the p7NS2 unit and the RC . The core protein and other viral proteins ( e . g . NS4B ) create the LD-ER microenvironment by redistribution of the LDs and intracellular membranes . The LDs surrounded by core protein are recruited to the RC . E1E2 complex interacts with p7NS2 unit and E1E2p7NS2 arrives to NS5A positive membranes in the proximity of LDs due to a combination of signals in p7 , NS2 and E2 proteins . NS5A switches from the replication to assembly mode by phosphorylation , which stabilizes the presence of NS2 in dotted structures favoring the assembly process ( Figure 10 ) . Finally , our data indicate a crucial role played by NS2 in the assembly process and highlight the complexity of the mechanism of its action . In conclusion , NS2 emerges as an essential mediator between the structural and non-structural proteins in HCV assembly process .
293T human embryo kidney cells ( HEK293T ) , U2OS human osteosarcoma cells ( American Type Culture Collection ) and Huh-7 human hepatoma cells [55] were grown in Dulbecco's modified essential medium ( Invitrogen ) supplemented with 10% fetal bovine serum . Anti-HCV Mabs A4 ( anti-E1 ) [56] and 3/11 ( anti-E2; kindly provided by J . A . McKeating , University of Birmingham , UK ) [57] were produced in vitro by using a MiniPerm apparatus ( Heraeus ) as recommended by the manufacturer . Anti E2 Mab AP33 was kindly provided by A . H . Patel , University of Glasgow , UK . Anti-capsid ACAP27 [58] and anti-NS3 ( 486D39 ) Mabs were kindly provided by JF Delagneau ( Bio-Rad , France ) . The anti-NS5A Mab 9E10 [8] and polyclonal antibody were kindly provided by CM Rice ( Rockefeller University , NY , USA ) and M Harris ( University of Leeds , UK ) , respectively . The anti-NS2 Mab 6H6 [18] and polyclonal antibody were kindly provided by CM Rice ( Rockefeller University , NY , USA ) and R Bartenschlager ( University of Heidelberg , Germany ) , respectively . The anti-Sec31 antibody [38] was kindly provided by F Gorelick ( Yale University School of Medicine , CT , USA ) . The anti-p125 Mab [37] was kindly provided by K Tani ( University of Tokyo , Japan ) . The following antibodies were purchased: the anti-ERGIC-53 Mab ( Alexis ) , the anti-actin ( Santa Cruz Biotechnology ) , anti-calnexin polyclonal ( Stressgen ) , anti-calreticulin polyclonal ( Stressgen ) , anti-PDI ( Stressgen ) , anti-GFP ( Roche ) and the anti-hemagglutinin ( HA ) Mab 3F10 ( Roche ) and Mab HA11 ( Covance ) . Alexa 488 , Alexa 555 and Alexa 633 conjugated goat anti–rabbit , anti-rat and anti-mouse immunoglobulin G ( IgG ) were purchased from Invitrogen . All plasmids were constructed by standard molecular biology methods and the constructs were confirmed by sequencing . The following plasmids were assembled for FRET-FLIM analyses in the background of pCMV plasmid ( Addgene ) : pCMV/YFP-E1TM , pCMV/CFP-E2TM , pCMV/CFP-p7 , pCMV/CFP-EYF , pCMV/YFP-NS2 . For these constructs , the plasmids encoding YFP and CFP were obtained from DW Piston ( Vanderbilt University , USA ) and A Miyawaki ( Riken Institute , Japan ) , respectively . YFP-E1TM and CFP-E2TM are YFP and CFP in fusion with the transmembrane domain of E1 and E2 ( H strain ) , respectively . In these two constructs , a two amino acid linker ( serine and glycine ) was inserted between the fluorescent protein and the transmembrane domains . CFP-p7 and CFP-EYF are CFP proteins fused to the N-terminus of p7 and the transmembrane domain of yellow fever virus envelope protein E ( EYF ) , respectively . YFP-NS2 is a YFP protein fused to the N-terminus of NS2 . In all the constructs , the calreticulin signal sequence was fused to the N-terminus of CFP and YFP for translocation in the ER lumen , allowing the study of the recombinant proteins in their native topology . For p7-NS2 co-immunoprecipitation experiments , the following plasmids were constructed in the background of pTriex ( Novagen ) or pCI plasmids ( Promega ) : pTriex/p7-Flag , pTriex/EYF-Flag , pCI/HA-NS2 , pCI/HA-NS2-GFP and pCI/HA-NS2-GTM . p7-Flag and EYF-Flag are HCV p7 of genotype 1a ( H strain ) and YFE in fusion with the Flag epitope ( DYKDDDDK ) . HA-NS2 has a HA epitope fused to the N-terminus of NS2 . In HA-NS2-GFP , the cytosolic domain of NS2 was replaced by GFP protein , whereas for HA-NS2-GTM , we replaced the transmembrane domain of NS2 by the transmembrane domain of VSV-G protein . In this work , we used a modified version of the plasmid encoding JFH1 genome ( genotype 2a; GenBank access number AB237837 ) , kindly provided by T . Wakita ( National Institute of Infectious Diseases , Tokyo , Japan ) [7] . Mutations were introduced in a JFH1 plasmid containing a Renilla Luciferase reporter gene [59] and mutations leading to amino acids changes F172C and P173S which have been shown to increase the viral titers [30] . Furthermore , the E1 sequence encoding residues 196TSSSYMVTNDC has been modified to reconstitute the A4 epitope ( SSGLYHVTNDC ) [56] as described [31] . Overlapping PCR was used to construct all the JFH1 mutants . The JFHGND-Luc construct was obtained by inserting the previously described GND mutation [7] in JFH-Luc plasmid . The JFH-HA construct was obtained by inserting the sequence of the HA epitope ( YPYDVPDYA ) followed by a GGG linker at the N-terminus of NS2 . The JFH-Δp7-HA keeps the first 2 amino acids of p7 followed by the HA tag sequence and the SGG linker at the N-terminus of NS2 . The JFH-ΔE1E2-Luc plasmid has been described previously [31] . It contains an in-frame deletion of amino acids 217–567 . JFH-ΔTM12-HA has the first two transmembrane segments of NS2 deleted as described [19] , in the context of our JFH-HA virus . JFH-HA-PP has proline residues 138 and 143 in domain D2 of the core protein replaced by alanine residues as described [13] , [41] . JFH-S/A-HA has serine 457 of NS5A replaced by an alanine as described [17] . JFH-3BS/A-HA has serine residues at positions 452 , 454 and 457 of NS5A replaced by alanine residues as described [16] . The JFH-S/D-HA has serine 457 of NS5A replaced by an aspartate residue as described [17] . JFH-ΔTME2-HA and JFH-ΔE1E2TME2-HA have the transmembrane region of E2 glycoprotein replaced by FMDV 2A autoprotease . JFH-ΔTME2-HA contains an in-frame deletion of amino acids 720–750 , corresponding to the transmembrane domain of E2 of JFH1 , whereas JFH-ΔE1E2TME2-HA contains in-frame deletions of amino acids 217–567 and 720–750 . To construct the JFH-ΔTME2-HA , we replaced the C-terminal region of E2 ( aa 720–750 of JFH1 ) by QLLNFDLLKLAGDVESNPGP FMDV 2A autoprotease peptide preceded by a GGG linker sequence . A similar strategy was used for the construction of JFH-ΔE1E2TME2-HA using as a backbone plasmid the JFH-ΔE1E2-HA . JFH-RR/QQ-HA has the arginines 33 and 35 of p7 replaced by glutamine residues . The primers and enzymes used for the constructs are presented in Table S1 . Schematic representation of the constructs used in this study is presented in Figure S1 . Twenty-four hours before transfection , 293T cells were seeded in 100 mm tissue culture plates to reach a 70–80% confluency the next day . Cells were tranfected with 6 µg of DNA/plate at a ratio of 1∶4 with PEI transfection reagent ( Eurogentec ) . In cotransfection experiments , equal quantities of each plasmid were used . At 24h post-transfection , cells were processed for co-immunoprecipitation analyses . Twenty-four hours before transfection , U2OS cells were seeded in 6 well plates on 32mm slides to reach a confluency of 70–80% the next day . Cells were transfected with 1µg of CFP-expressing plasmid ( donor ) and 125 ng of YFP-expressing plasmid ( acceptor ) mixed with Fugene reagent ( Roche ) following the instructions of the manufacturer . For CFP-E1 and YFP-E2 co-transfection experiments , we used 300 ng of donor and 600 ng of acceptor plasmids , respectively . Twenty-four hours after transfection , U2OS cells were selected for FRET-FLIM acquisition . We analyzed cells with similar expression levels of donor and acceptor fusion proteins . We also chose cells with normal reticulate ER morphology avoiding those where the overexpression of recombinant proteins was present . In order to detect the FRET events , the Time Correlated Single Photon Counting FLIM system ( TCSPC ) was used [60] , [61] , [62] . The analyses were performed with a Leica SP5 . X confocal Microscope ( Leica Microsystem ) with an internal FLIM detector . A dedicated photo-counting and timing electronic card ( SPC 830 TCSPC card , Becker and Hickl ) was coupled to the Leica internal detector and used to classify the photon emission in time to determine the lifetime of the donor protein . To excite the samples , Chameleon Ultra2 ( Chorent Inc ) biphoton was used at 830 nm at an average power of 0 . 13 mW/µm2 [60] , [61] , [62] . The fluorescence events of the donor protein result in a photon decay curve generated by the FLIM method . The decay curve was directly used to determine the donor's lifetime . The least square fitting method was used to describe the non linear responses commonly observed in FLIM analysis [60] , [61] , [62] . We used TITAN ( “in the house” designed ) and SPCImage ( Becker and Hickl ) software for advanced FLIM data analysis and curve fittings . In order to reduce the impact of background and improve the Signal to Noise Ratio ( SNR ) , we excluded from our analysis the pixels located in the nuclear region or from ER-like regions where the donor protein was overexpressed . After setting these thresholds , we made a summation of all the pixels of interest to achieve a fitting statistically significant for the TITAN software . To compensate the possible large scattering of points in the curve , we used a Newton trust region algorithm [63] and an extraction of mean lifetimes was performed in order to determine the FRET events from the multi-exponential model [64] . Cell pellets were lysed in phosphate-buffered saline ( PBS ) lysis buffer ( 1% Triton 100-X , 20mM NEM , 2mM EDTA , protease inhibitors cocktail Roche ) and they were precleared with 20 µl Prot G for 2h at 4C . The precleared lysates were incubated with anti-HA antibodies ( HA11 ) or Sepharose beads covalently bound to HA11 antibody ( Covance ) overnight at 4h . The immunocomplexes were pulled down with 50 µl of Protein G and washed three times with lysis buffer . For p7-NS2 interaction , cell lysates were incubated with 20 µl of agarose-anti-Flag beads over night at 4°C . The immunocomplexes were treated the same as above and the Western blots were revealed by an anti-HA antibody . The endoglycosidase digestions were performed following the manufacture's instructions ( NEB ) . Briefly , cell lysates containing 20 µg of protein were denatured in EndoH ( PNGase ) denaturing buffer ( 0 . 5% SDS , 1% 2-mercaptoethanol ) for 10 min at 100°C . Then , the lysates were incubated or not with 1 µl of EndoH ( PNgase ) for 20h at 37°C . After separation by SDS-PAGE , proteins were transferred to nitrocellulose membranes ( Hybond-ECL , Amersham ) by using a Trans-Blot apparatus ( Biorad ) and revealed with specific antibodies followed by secondary immunoglobulin conjugated to peroxidase . The proteins of interest were revealed by enhanced chemiluminescence detection ( ECL , Amersham ) as recommended by the manufacturer . Plasmids encoding wild-type ( WT ) and mutated genomes were linearized at the 3′ end of the HCV cDNA with the restriction enzyme XbaI and treated with the Mung Bean Nuclease ( New England Biolabs ) . In vitro transcripts were generated using the Megascript kit according to the manufacturer's protocol ( Ambion ) . The in vitro reaction was set up and incubated at 37°C for 4 h and transcripts were precipitated by the addition of LiCl . Ten micrograms of RNA were delivered into Huh-7 cells by electroporation as described previously [30] . Replication was assessed at 72 h by measuring Renilla Luciferase activities in electroporated cells as indicated by the manufacturer ( Promega ) . Supernatants containing HCVcc were harvested 72 h after electroporation and filtered through 0 . 45 µm pore-sized membrane for infectivity measurements . HCVcc were incubated for 3 h with Huh-7 cells seeded the day before in 24-well plates . At 72 h post-infection , Luciferase assays were performed on infected cells as indicated by the manufacturer ( Promega ) . For supernatants titration , Huh7 electroporated cells were seeded in 6-well plates . 72h post-electroporation , naïve Huh-7 cells were inoculated with serial dilutions of the supernatant . 48h post-inoculation , the infected cells were fixed in ice-cold methanol , they were immunostained with anti-E1 antibody and the focus forming units ( FFUs ) were counted . HCV Core was quantified by a fully automated chemiluminescent microparticle immunoassay according to manufacturer's instructions ( Architect HCVAg , Abbott , Germany ) [65] , [66] . For the determination of intracellular core quantity , the electroporated cells were lysed in PBS lysis buffer ( 1% Triton 100-X , 20mM NEM , 2mM EDTA , protease inhibitors cocktail Roche ) and the lysates were cleared for 20 min at 14 , 000g . Mutated HCV genomes were delivered into Huh-7 cells . At day 3 post-electroporation , cells were trypsinized , washed once with fresh medium and reseeded into cell culture dishes . At day 5 post-electroporation , total RNA in cell lysates and HCV RNA in supernatants were extracted using the Qiagen RNeasy kit and Qiagen QiaAmp viral RNA mini kit , respectively . cDNA was synthesized using High Capacity cDNA Reverse Transcription kit as described by the manufacturer ( Applied BioSystems ) and titrated by quantitative real-time RT-PCR assay ( RT-qPCR ) using TaqMan and minor groove binding ( MGB ) probe detection . The primer pair and the probe were located in the 5′ HCV non-coding region [67] . Huh-7 cells transfected with HCV RNA were grown on 12-mm glass coverslips . At the indicated time points , cells were fixed with 3% paraformaldehyde and then permeabilized with 0 . 1% Triton X-100 in PBS . Both primary- and secondary-antibody incubations were carried out for 30 min at room temperature with PBS containing 10% goat serum . LDs were stained for 10 minutes in 300 ng/ml BODIPY 493/503 ( Invitrogen ) . Nuclei were stained with 4 , 6′-diamidino-2-phenylindole ( DAPI ) . The coverslips were mounted on slides by using Mowiol 4–88 ( Calbiochem ) containing mounting medium . Confocal microscopy was performed with an LSM710 laser-scanning confocal microscope ( Zeiss ) using a 63×/1 . 4 numerical aperture oil immersion objective . Signals were sequentially collected by using single fluorescence excitation and acquisition settings to avoid crossover . Images were processed using Adobe Photoshop software . Cells showing NS2/NS5A-positive dot-like structures were counted on images collected with a 40× oil immersion objective . Huh-7 cells transfected with HCV RNA were grown on 75 cm2 flasks . At 72h post-electroporation , cells were fixed by incubation in a solution containing 4% paraformaldehyde in 0 . 1 M phosphate buffer ( pH 7 . 2 ) for 20 h . The cells were collected by centrifugation and the cell pellet was then dehydrated in a graded series of ethanol solutions at −20°C , using an automatic freezing substitution system ( AFS , Leica ) , and embedded in London Resin Gold ( LR Gold , Electron Microscopy Science ) . The resin was allowed to polymerize at −25°C , under UV light , for 72 h . Ultrathin sections were cut and blocked by incubation with 3% fraction V bovine serum albumin ( BSA , Sigma ) in PBS . They were then incubated with anti-HA Mab ( Covance ) diluted 1∶50 in PBS supplemented with 1% BSA , washed and incubated with goat anti-mouse antibodies conjugated to 15 nm gold particles ( British Biocell International , Cardiff , UK ) diluted 1∶40 in PBS supplemented with 1% BSA . Ultrathin sections were cut , stained with 5% uranyl acetate , 5% lead citrate , and placed on electron microscopy grids coated with collodion . The sections were then observed with a Jeol 1230 transmission electron microscope ( Tokyo , Japan ) connected to a Gatan digital camera driven by Digital Micrograph software ( Gatan , Pleasanton , CA ) for image acquisition . | Hepatitis C virus ( HCV ) causes major health problems worldwide . Understanding the major steps of the life cycle of this virus is essential to developing new and more efficient antiviral molecules . Virus assembly is the least understood step of the HCV life cycle . Growing experimental evidence indicates that , in addition to the physical virion components , the HCV non-structural proteins are intimately involved in orchestrating morphogenesis . Since it is dispensable for HCV RNA replication , the non-structural viral protein NS2 is suggested to play a central role in HCV particle assembly . Molecular interactions between NS2 and other HCV proteins were demonstrated . Furthermore , NS2 was shown to accumulate over time in endoplasmic reticulum-derived structures and to colocalize with the viral envelope glycoproteins and viral components of the replication complex in close proximity to the HCV core protein and lipid droplets . Importantly , specific mutations within NS2 that affected HCV infectivity could also alter the subcellular localization of NS2 protein and its interactions , suggesting that this subcellular localization and its interactions are essential for HCV particle assembly . Altogether , these observations indicate that NS2 protein plays an important role in connecting different viral components that are essential for virus assembly . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"virology/virion",
"structure,",
"assembly,",
"and",
"egress",
"virology"
]
| 2011 | NS2 Protein of Hepatitis C Virus Interacts with Structural and Non-Structural Proteins towards Virus Assembly |
Cyclin A is critical for cellular DNA synthesis and S phase progression of the cell cycle . Human cytomegalovirus ( HCMV ) can reduce cyclin A levels and block cellular DNA synthesis , and cyclin A overexpression can repress HCMV replication . This interaction has only been previously observed in HCMV as murine CMV does not downregulate cyclin A , and the responsible viral factor has not been identified . We previously reported that the HCMV protein pUL21a disrupted the anaphase-promoting complex ( APC ) , but a point mutant abrogating this activity did not phenocopy a UL21a-deficient virus , suggesting that pUL21a has an additional function . Here we identified a conserved arginine-x-leucine ( RxL ) cyclin-binding domain within pUL21a , which allowed pUL21a to interact with cyclin A and target it for proteasome degradation . Homologous pUL21a proteins from both chimpanzee and rhesus CMVs also contained the RxL domain and similarly degraded cyclin A , indicating that this function is conserved in primate CMVs . The RxL point mutation disabled the virus' ability to block cellular DNA synthesis and resulted in a growth defect similar to pUL21a-deficient virus . Importantly , knockdown of cyclin A rescued growth of UL21a-deficient virus . Together , these data show that during evolution , the pUL21a family proteins of primate CMVs have acquired a cyclin-binding domain that targets cyclin A for degradation , thus neutralizing its restriction on virus replication . Finally , the combined proteasome-dependent degradation of pUL21a and its cellular targets suggests that pUL21a may act as a novel suicide protein , targeting its protein cargos for destruction .
Human cytomegalovirus ( HCMV ) , a widespread β-herpesvirus that establishes a lifelong infection , is capable of causing severe complications in immuno-naïve populations and immuno-compromised patients [1] . HCMV is known to use a number of mechanisms to manipulate the host cell cycle so that infected cells are arrested at the G1/S boundary . Some of these mechanisms are inhibition of Rb and activation of the E2F family of transcription factors [2]–[6] , modulation of the anaphase-promoting complex ( APC ) [7] , [8] , suppression of the mini-chromosome maintenance complex ( MCM ) [9] , and alteration of cyclin/cyclin-dependent kinase ( CDK ) activity [10]–[15] . It has been postulated that these regulations provide essential nutrients and cellular enzymes needed for viral DNA synthesis while preventing cellular DNA synthesis from competing for these important resources . Cyclins are regulatory proteins that interact with CDKs to phosphorylate numerous substrates involved in cell cycle progression . It has been established that HCMV dramatically affects the levels and activity of several cyclin-CDK complexes [5] , [12] , [15]–[18] , as well as CDK inhibitors such as p21 [13] , [19] , [20] . During infection , cyclins B and E are upregulated , cyclin D is largely unchanged , and cyclin A is inhibited [12] , [15]–[17] , whereas CDK levels are largely unaffected [15] . While the roles of cyclins B , E , and D on virus replication and virus-induced cell cycle manipulation remain to be determined , overexpression of cyclin A represses HCMV replication [14] . Cells overexpressing cyclin A had delayed IE viral gene expression , and a more noticeable block on splicing of IE transcripts as well as early and late gene expression , resulting in a multi-log reduction in viral titers . While it is likely that HCMV actively represses the expression of cyclin A to avoid its detrimental effects , the viral factor responsible and its mechanism of action remain unknown . Importantly , murine CMV ( MCMV ) does not actively block cyclin A expression or activity , and cyclin A overexpression does not affect MCMV gene expression [14] , suggesting that the antiviral activity of cyclin A is specific to human or primate CMVs . We have previously identified an HCMV protein , pUL21a , which is required for efficient virus growth by promoting viral DNA synthesis and late accumulation of spliced IE transcripts [21] , [22] . Recently we found that this protein disrupted the APC by proteasome-dependent degradation of subunits APC4 and APC5 [7] . However , a point mutation that rendered pUL21a unable to regulate the APC ( PR109-110AA ) was not sufficient to alter virus growth by itself . This pUL21 point-mutant virus was attenuated only when pUL97 , the second viral APC regulator , was simultaneously deleted , indicating that APC regulation is required for efficient HCMV replication . The different phenotypes between UL21a deletion virus and APC-binding point mutant virus also suggest that pUL21a must have an additional function . Here we show that pUL21a interacts with and targets cyclin A for proteasome-dependent degradation through its primate CMV conserved RxL cyclin-binding domain . This activity is completely independent of its ability to inhibit the APC . Loss of this domain results in a marked increase in cellular DNA replication and significantly attenuates virus replication . Furthermore , knockdown of cyclin A can rescue pUL21a mutant virus replication , indicating that pUL21a targets cyclin A for degradation to counter its restriction activity during HCMV replication .
Using the Eukaryotic Linear Motif Resource ( ELM ) [23] , we identified an arginine-x-leucine ( RxL ) cyclin-binding domain [24] , [25] in pUL21a proteins from human CMV ( HCMV ) , chimpanzee CMV ( ChCMV ) , and rhesus CMV ( RhCMV ) ( Fig . 1A ) . To determine whether pUL21a was able to bind to cyclins , we transfected 293T cells with constructs expressing GFP-tagged pUL21a isoforms and immunoprecipitated cell lysates with GFP antibody ( Fig . 1B ) . Cyclin A is present in two forms: cyclin A1 that is expressed during embryogenesis and cyclin A2 that is expressed in all other dividing cells . We found that GFP-tagged pUL21a co-immunoprecipitated with cyclin A2 ( termed cyclin A herein ) but not cyclins B or E ( lane 5 ) . This interaction was independent of GFP as the tagged UL21a stop mutant ( gfpUL21astop ) [7] failed to co-immunoprecipitate with any cyclins ( lane 6 ) . Moreover , this interaction was also independent of the binding of pUL21a to the APC , as the APC-binding mutant ( gfpUL21aPR-AA ) [7] co-immunoprecipitated with cyclin A even though it failed to pull down APC3 ( lane 7 ) . Importantly , mutation of the RxL domain ( gfpUL21aRxL-AxA ) abrogated pUL21a's ability to co-immunoprecipitate with cyclin A but not APC3 ( lane 8 ) , indicating that the RxL domain is necessary and specific for pUL21a to bind to cyclin A but not the APC . These interactions were also confirmed by reciprocal immunoprecipitation using cyclin A antibody . As expected , cyclin A co-immunoprecipitated with wildtype and PR-AA mutant pUL21a ( lanes 9 and 11 ) , but not the stop mutant or RxL mutant pUL21a ( lanes 10 and 12 ) . To test whether pUL21a interacts with cyclin A during HCMV infection , we performed an immunoprecipitation assay on cell lysates infected with control wildtype virus expressing free GFP ( ADgfp ) or recombinant HCMV expressing GFP-tagged UL21a ( ADgfpUL21a ) ( Fig . 1C ) . GFP tagged pUL21a co-immunoprecipitated with cyclin A and APC3 but not cyclins B and E ( lane 4 ) , and cyclin A co-immunoprecipitated with both native and GFP tagged pUL21a ( lanes 5 and 6 ) . Together , these results indicate that pUL21a uses its conserved RxL cyclin-binding domain to interact with cyclin A , and that the ability of pUL21a to bind to cyclin A or the APC is independent of each other . To test the consequence of pUL21a binding on cyclin A protein levels , we analyzed cyclin A accumulation during infection of recombinant HCMV viruses . Fibroblasts were synchronized by contact inhibition and released 24 hours before infection . This results in a sharp rise and then decrease of cyclin A levels over 72 hours . Consistent with previous studies [15] , [16] , wildtype virus ( ADgfp ) markedly reduced cyclin A protein levels during infection ( Fig . 2A ) . However , cells infected with UL21a-deficient virus ( ADsubUL21a ) or cyclin binding-deficient virus ( ADpmUL21aRxL-AxA ) accumulated much higher levels of cyclin A . The inability of ADpmUL21aRxL-AxA virus to reduce cyclin A accumulation was not due to an inability to induce APC subunit degradation as it degraded APC4 as efficiently as wildtype virus ( Figs . 2A and 2B ) . This inability was also not the result of the somewhat reduced stability of mutant pUL21a at 72 hours post infection ( hpi ) , as the mutant virus could degrade the APC efficiently at this time point . Furthermore , mutant virus ADpmUL21aPR-AA , which was unable to inhibit the APC , was still able to reduce cyclin A protein levels ( Fig . 2B ) . Thus , not only the binding ( Fig . 1 ) , but also the ability to reduce cyclin A and APC subunit protein levels by pUL21a are genetically separable functions that do not depend on each other . Differences in cyclin A levels between wildtype and UL21a-deficient virus were apparent as early as 6–9 hours post infection ( hpi ) and coincides with expression kinetics of pUL21a ( Fig . 2C ) . This suggests that the phenotype is not due to reduced levels of late gene expression seen in UL21a-deficient virus infection [22] . Together , these results indicate that pUL21a is necessary for virus-induced cyclin A reduction during HCMV infection . To test if pUL21a was sufficient to reduce cyclin A protein levels , we used a tetracycline-inducible system to over-express pUL21a in MRC-5 cells [7] . Tetracycline induced pUL21a expression , leading to the loss of cyclin A expression when compared to the non-induced control ( Fig . 2D ) . While both wildtype and RxL mutant pUL21a reduced APC4 levels , mutant pUL21a failed to reduce cyclin A levels , further supporting the conclusion that pUL21a regulation of cyclin A but not APC4 is dependent on the RxL cyclin-binding domain . As the RxL domain is present in UL21a proteins of primate CMVs ( Fig . 1A ) , we hypothesized that cyclin A regulation is a conserved function of pUL21a . To test this , we used the same inducible system to express UL21a proteins from ChCMV and RhCMV . Both viral proteins were able to reduce cyclin A levels ( Fig . 3A ) . Furthermore , both proteins were able to rescue UL21a-deficient HCMV virus to levels similar to that by HCMV pUL21a ( Fig . 3B ) . Importantly , the RhCMV UL21a protein did not degrade APC4 ( Fig . 3A ) but still rescued UL21a-deficient HCMV virus ( Fig . 3B ) , providing evidence that the activity of pUL21a to inhibit cyclin A is critical for efficient HCMV replication . The ChCMV UL21a-expressing cells overall did not support HCMV infection as well as other cells did , likely due to differences in the nature of the protein and where in the cellular genome lentivirus was transduced . Together , these data strongly suggest that pUL21a-mediated reduction of cyclin A is a conserved and important function of primate CMVs . We next investigated the potential mechanism of how pUL21a reduced cyclin A levels . pUL21a became detectable at 10 hpi and peaked at 24–48 hpi ( Fig . 2A ) [21] . Its protein levels markedly increased in the presence of proteasome inhibitors , indicating that this viral protein was inherently unstable and was targeted for proteasome degradation during HCMV infection [21] . Intriguingly , pUL21a also induced proteasome-mediated degradation of APC4 and APC5 [7] . We therefore hypothesized that pUL21a regulated cyclin A levels by inducing its proteasome-dependent degradation . To test this , we measured cyclin A protein levels in the presence of the proteasome inhibitors MG132 and epoxomicin during infection . Both MG132 and epoxomicin restored cyclin A protein levels in wildtype infection to those seen in mock and UL21a-deficient virus infected cells ( Fig . 4A ) . While cyclin A transcript levels were reduced during infection as previously reported [12] , [16] , wildtype and UL21a-deficient virus had similar levels of cyclin A transcripts ( Fig . 4B ) . Moreover , cyclin A protein levels remained high despite drastic reduction in its transcript levels following MG132 treatment ( Figs . 4A and 4B ) . These data suggest that regulation of cyclin A by pUL21a occurs at the protein levels . Cyclin A contains a D-box motif that is necessary for recognition by the APC , the cellular E3 ubiquitin ligase that ubiquitinates and targets cyclin A for degradation during the cell cycle [14] , [26] , [27] . To test if the D-box motif was required for pUL21a-mediated degradation of cyclin A , we overexpressed tetracycline-inducible FLAG-tagged wildtype and D-box mutant cyclin A by lentiviral transduction in HCMV-infected cells . In the absence of tetracycline , only endogenous cyclin A was expressed , and its protein levels were reduced by almost 8-fold during wildtype virus infection as compared to those during pUL21a-deficient virus infection ( Fig . 4C ) . In the presence of tetracycline , expressions of FLAG-tagged cyclin A was induced , and their expression was driven by the CMV promoter that was further potentiated during CMV infection . Therefore , unlike endogenous cyclin A , the levels of FLAG-tagged cyclin As were higher in HCMV-infected cells relative to those in mock-infected cells ( Fig . 4C ) . Nonetheless , in the presence of tetracycline , pUL21a was able to similarly reduce protein levels of endogenous cyclin A , overexpressed FLAG-tagged wildtype , and D-box mutant cyclin A during HCMV infection ( Fig . 4C ) . We noted that cyclin A reduction by pUL21a in the presence of tetracycline was less pronounced than that in the absence of tetracycline ( i . e . ∼8-fold vs . ∼3-fold ) ( Fig . 4C ) . The lesser reduction of cyclin A in the presence of tetracycline was likely due to overexpression of cyclin A which overwhelmed the capacity of pUL21a to target it for degradation . Overall , the ability of pUL21a to degrade D-box cyclin A mutant , along with independent binding and degradation of the APC and cyclin A by pUL21a ( Figs . 1 and 2 ) , suggest that pUL21a-mediated degradation of cyclin A is independent of the APC . Together , our data show that pUL21a regulates cyclin A by inducing its proteasome-dependent degradation , and that this regulation is likely independent of the APC , the cellular E3 ligase known to regulate cyclin A ubiquitination and degradation . HCMV inhibits cellular DNA synthesis [28] , and cyclin A is a critical factor for promoting cellular DNA synthesis . We hypothesized that pUL21a regulation of cyclin A was important to prevent cellular DNA synthesis during HCMV infection . To test this , we infected MRC-5 fibroblasts with recombinant virus and measured cellular DNA synthesis by flow cytometry . Cells were gated by pUL44 expression to distinguish infected cells from uninfected cells ( Fig . 5A ) . No differences in cellular DNA content were seen at 24 hpi ( Fig . 5B ) . However by 48 hpi , while wildtype and APC-binding mutant viruses continued to inhibit cellular DNA synthesis , allowing only 12–14% of cells to enter S phase , UL21a-deficient and RxL mutant viruses failed to do so , resulting in nearly twice as many cells in S phase ( 25–28% ) ( Figs . 5B–C ) . We conclude that pUL21a-mediated cyclin A degradation is one mechanism used by HCMV to block cellular DNA synthesis during infection . In the final experiments , we tested the consequence of pUL21a-mediated cyclin A degradation on HCMV replication in fibroblasts . It was previously shown that cyclin A overexpression inhibited HCMV replication by several logs and was responsible for the block in lytic gene expression during the S/G2 phases [13] , [14] . Consistent with this , UL21a-deficient and RxL mutant viruses were also attenuated , with a growth defect of ∼3 logs relative to wildtype and marker rescued virus ( Fig . 6A ) . In addition , marker rescued virus reduced cyclin A protein levels as efficiently as wildtype virus , indicating that the defect in cyclin A regulation during RxL mutant virus infection is not due to second-site mutations ( Fig . 6B ) . These data together suggest that cyclin A degradation by pUL21a is necessary for effective virus replication . To provide additional proof testing this hypothesis , we knocked down cyclin A in MRC-5 fibroblasts and infected these cells with wildtype and UL21a-deficient virus . Cyclin A knockdown rescued virus production and IE2 expression ( IE2-86 , IE2-60 , and IE2-40 ) of UL21a-deficient virus to levels equivalent to wildtype virus ( Figs . 6C–D ) . Wildtype virus showed slightly reduced growth in the presence of cyclin A knockdown , likely reflecting an off-target effect of this siRNA or a consequence of the loss of an essential host protein . Together , these data suggest that cyclin A degradation by pUL21a is necessary for effective virus replication . We conclude that pUL21a antagonizes the cyclin A-mediated restriction on HCMV infection by binding to and promoting proteasome degradation of this prominent cell cycle regulator .
Here we describe a novel function for the HCMV protein pUL21a to bind to the important cell cycle regulator cyclin A and direct it for proteasome degradation , which represents a novel viral mechanism to combat this host factor . This interaction is mediated by a highly conserved RxL cyclin-binding domain that is present in pUL21a proteins of all primate CMVs tested in this study . pUL21a from ChCMV and RhCMV not only similarly reduce cyclin A levels , they also rescued the growth defect of pUL21a-deficient HCMV virus . pUL21a is a highly unstable protein subject to rapid proteasome-mediated degradation [21] . Intriguingly , it binds to two major cell cycle regulators , the anaphase-promoting complex ( APC ) [7] and cyclin A , and targets both for degradation in a proteasome-dependent manner . Finally , this pUL21a-mediated cyclin A degradation bears an important consequence on virus replication . Mutation in the RxL cyclin-binding domain of pUL21a destroys the ability of HCMV to inhibit cyclin A , resulting in unchecked cellular DNA synthesis and severe attenuation of virus growth . Importantly , growth of the UL21a-deficient virus can be rescued by knocking down cyclin A , providing the definitive proof that pUL21a targets cyclin A for degradation in order to antagonize its innate restriction on HCMV replication . Our work suggests that pUL21a exploits the proteasome pathway in a way that it independently regulates APC and cyclin A protein levels . We have recently reported that pUL21a binds to and targets APC subunits 4 and 5 for degradation , leading to the dissociation of this complex [7] . Now we show that a pUL21a point mutant unable to target APC4 and APC5 ( i . e . PR-AA mutation ) is still able to target cyclin A for degradation , and conversely , a point mutant unable to target cyclin A ( RxL-AxA ) is still able to target APC4 for degradation . We conclude that these two functions of pUL21a are independent of each other and propose the model where pUL21a uses two distinct domains to uniquely interact with these cell cycle regulators ( Fig . 7 ) . Whether pUL21a uses separate or similar mechanisms to target them for proteasome degradation remains to be determined . Regardless , these two functions of pUL21a lead to distinct phenotypic effects ( Fig . 7 ) . It has been known that HCMV promotes cells into a pseudo G1/S phase by inducing expression of S-phase genes but simultaneously blocking cellular DNA synthesis . pUL21a-induced degradation of APC subunits , along with pUL97-mediated phosphorylation of the APC co-activator Cdh1 [8] , leads to increased levels of APC substrates , which would promote entry into S phase [29] . However , inhibition of the APC could also increase cyclin A levels , which is detrimental to HCMV replication . Therefore , HCMV has developed a second function within pUL21a , which is to promote cyclin A degradation . This will antagonize the antiviral activity of cyclin A , inhibit cellular DNA synthesis , and phenotypically prevent infected cells from entering S phase . Thus the virus has elegantly evolved two mechanisms within one protein to harness the benefits of inhibiting the APC as well as overcoming any detrimental consequences of such regulation . How does pUL21a target cyclin A for proteasome degradation ? Cyclin A is normally targeted for ubiquitination and proteasome degradation during M phase of the cell cycle . Preliminary work suggests a modest increase in cyclin A ubiquitination in the presence of pUL21a ( data not shown ) , so it is possible that pUL21a may have a role in cyclin A ubiquitination . If so , as pUL21a contains no domain indicating it as an E3 ligase , it could recruit an E3 ligase to ubiquitinate cyclin A or block the activity of a deubiquitinase ( DUB ) . Alternatively , it is also entirely possible that pUL21a may directly target cyclin A to the proteasome in a ubiquitin-independent manner . Cyclin A is normally targeted by the APC for ubiquitination and proteasome degradation . However , inhibition of the APC by pUL21a and pUL97 [7] , [8] , along with our present data showing that pUL21a can mediate cyclin A degradation in the absence of its APC recognition motif ( D-box ) , suggests that cyclin A degradation during HCMV infection is APC-independent . Interestingly , pUL21a itself is rapidly degraded by the proteasome in a ubiquitin-independent manner , and is highly sensitive to proteasome inhibition [21] . It is tempting to speculate that pUL21a may act as a novel “suicide” protein that delivers substrates to the proteasome and is degraded with them . Mechanistically , pUL21a could directly bind to the proteasome , or could be recognized as an unfolded protein due to its proline-rich , predicted unstructured C-terminus tail [7] , and transported to the proteasome by chaperones . Additional biochemical , genetic , and structural analysis will be critical to testing these hypotheses . What is the mechanism for cyclin A to restrict virus replication ? Cyclin A and its concomitant CDK's phosphorylate a number of substrates that promote cellular DNA synthesis . It is possible that active cellular DNA synthesis would limit the resources available for viral genome amplification and interfere with the process of viral DNA replication , ultimately attenuating virus growth . Thus , pUL21a mutants may provide us with a tool to test the long held belief that cellular DNA replication is detrimental to herpesvirus replication . Alternatively , a substrate or set of substrates of cyclin A may also have more direct and specific antiviral activity that blocks HCMV replication . This pUL21a-mediated cyclin A regulation appears to be unique to CMVs of high mammals , such as primates , as murine CMV does not encode a UL21a homologous protein . It is possible that murine CMV has evolved alternative mechanisms to arrest the cell cycle , or that its shorter life cycle ( ∼32 hours ) circumvents the need to inhibit cyclin A activity of host cells . Consistent with this , murine CMV upregulates cyclin A levels and initial viral gene expression is unaffected by cyclin A overexpression [14] , highlighting an interesting and striking difference between closely related virus families . Finally , the discovery of pUL21a as a cyclin A modulator can also provide a useful tool to delineate the activity of this important molecule in cell biology . Only recently have large-scale screenings been used to systematically identify cyclin A substrates in mammalian cells [30] . pUL21a may be used to confirm these recently identified cyclin A substrates , explore the role of these substrates in cellular DNA synthesis , or discover additional substrates that play roles in cellular DNA synthesis and proliferation . Viral systems have been instrumental in many seminal discoveries in the history of cell biology . Novel virus-encoded regulators such as pUL21a can be powerful tools to probe the biology of host cells that are otherwise difficult to study .
Human foreskin fibroblasts ( HFFs ) , primary embryonic lung fibroblasts ( MRC-5 ) , and 293T cells were propagated in Dulbecco's modified Eagle medium ( DMEM ) containing 10% fetal bovine serum and penicillin-streptomycin . Expression constructs were transfected into 293T cells with 1 mg/ml polyethyleneimine ( PEI , Polysciences ) in Opti-MEM media . Primers used in this study are listed in Table 1 . Point mutant UL21a sequences were constructed using PCR with the desired mutations incorporated into complementary primers . To create expression constructs , the RxL mutant UL21a PCR fragment was digested with restriction enzymes Bgl II and EcoR I , and ligated into a pLPCX ( Clontech ) -derived over-expression vector with an N-terminal GFP tag to create retroviral expression construct pYD-C760; or digested with Sal I and EcoR I and ligated into pYD-C639 , a pLKO-based lentiviral vector under a tetracycline-inducible CMV-TetO2 promoter [31] ( generous gift from Roger Everett , University of Glasgow ) , to create lentiviral expression construct pYD-C762 . RhCMV UL21a was amplified from RhCMV BAC [32] , ChCMV UL21a was constructed from long overlapping primers , and FLAG tagged wildtype cyclin A as well as D-box mutant cyclin A ( cyclin A ΔD ) were amplified from plasmids provided by Anindya Dutta ( University of Virginia School of Medicine ) [26] . They were all cloned into pYD-C639 to create lentiviral expression constructs . All other constructs used in this study have been previously described [7] . Lentivirus was produced by PEI transfection of corresponding expression constructs as described above , along with appropriate packaging plasmids , into 293T cells . To create stable expression cells , MRC-5 cells expressing GFP-TetR [7] were transduced with lentivirus , and selected with 2 µg/µL puromycin ( Sigma-Aldrich ) to produce stable cells expressing various forms of UL21a , FLAG-cyclin , or HA-ubiquitin under the CMV-TetO2 promoter . To knock down cyclin A by RNAi , MRC-5 cells were transfected with siGENOME siRNA against cyclin A or luciferase control ( Thermo Scientific ) , using the procedure of siLENTFECT ( Bio-Rad ) according to manufacturer's instructions . Cells were left in serum-free medium and infected 48 hours later in serum-containing medium . Primary antibodies used included anti-β actin ( AC-15 , Abcam ) ; anti-GFP ( 3E6 and A6455 , Invitrogen ) ; anti-APC3 ( 35/CDC27 , BD Biosciences ) ; anti-APC4 ( A301-176A , Bethyl laboratories ) ; anti-UL21a [21]; anti-IE2 ( mAB8140 , Chemicon ) ; anti-UL44 ( 10D8 , virusys ) ; anti-Cyclin A ( B-8 and H-432 , Santa Cruz ) ; anti-cyclin B ( GNS1 , Thermo Scientific ) ; anti-cyclin E ( HE12 , BD Biosciences ) ; anti-FLAG ( F1804 , Sigma ) ; anti-HA ( 16B12 , Covance ) ; and anti-IE1 ( generous gift from Thomas Shenk , Princeton University ) . BAC-HCMV clones used in the present study were constructed using PCR-based two-step linear recombination as previously reported [33] . pADgfp , which carried the genome of HCMVAD169 strain and a simian virus 40 early promoter-driven GFP gene in place of the viral US4–US6 region , was used to produce wildtype virus ADgfp [34] . pADpmUL21aRxL-AxA , which carried the point mutation RxL42-44AxA in the UL21a coding sequence , and pADrevUL21aRxL-AxA , in which RxL42-44AxA mutation was subsequently repaired , were used to produce UL21a cyclin-binding domain point mutant virus ADpmUL21aRxL-AxA and its marker rescued virus ADrevUL21aRxL-AxA , respectively . These recombinant BAC clones were confirmed by PCR , restriction digest , and sequencing . All other recombinant BAC-HCMV clones used were described previously [7] . Recombinant HCMV AD169 viruses were reconstituted from transfection of corresponding BAC-HCMV clones as previously described [33] . Viral stocks were harvested from infected cell culture supernatant and concentrated by ultracentrifugation through 20% D-sorbitol . Virus titers were determined in duplicate by tissue culture infectious dose 50 ( TCID50 ) assay on HFFs . Relative viral genome numbers were determined by extracting virion DNA and performing real-time quantitative PCR ( qPCR ) with either a taqman probe and primers specific to viral gene UL54 , or with SYBR green and primers to UL32 [21] . Confluent MRC-5 cells were split , and 24 hours later infected with HCMV at an input genome number equivalent to that of 3 infectious units of wildtype virus/cell , unless otherwise indicated . Cells were inoculated for 1 hour with virus and then replenished with fresh media . For cell cycle profiling , cells were treated with phosphonoacetic acid ( PAA , 100 µg/ml , Sigma-Aldrich ) immediately after infection . For proteosome inhibition experiments , cells were treated with MG132 ( 20 µM , Santa Cruz ) or epoxomicin ( 40 µM , Santa Cruz ) . Virus titers in the supernatant of infected cultures were determined by TCID50 assay . For immunoprecipitation , 293T and MRC-5 cells were lysed in NP40 buffer ( 0 . 5% NP40 , 50 mM Tris-Cl pH 8 . 0 , 125 mM NaCl ) and EBC2 buffer ( 0 . 5% NP40 , 50 mM Tris-Cl pH 8 . 0 , 300 mM NaCl ) , respectively . Lysis buffers were supplemented with protease and phosphatase inhibitors ( Roche and Sigma-Aldrich , respectively ) . Mouse anti-GFP antibody ( 3E6 , Invitrogen ) or mouse anti-Cyclin A antibody ( B-8 , Santa Cruz ) was conjugated to protein A dynabeads ( Invitrogen ) with BS3 ( Thermo Scientific ) according to manufacturer's instructions . Cleared cell lysates were incubated with conjugated dynabeads by gentle rotation at 4°C . Beads were washed once with lysis buffer and twice with PBS , and bound proteins were eluted in reducing sample buffer ( 200 mM Tris pH 6 . 8 , 6% SDS , 12% β-mercaptoethanol , 18% glycerol ) by incubating at 55°C . For whole cell lysate control ( WCL ) , cell lysates were similarly mixed with reducing sample buffer and incubated at 90°C for 5 minutes . For immunoblotting , cells were lysed in reducing sample buffer containing protease and phosphatase inhibitors . Proteins from equivalent cell numbers were resolved on a SDS poly-acrylamide gel , transferred to a polyvinylidene difluoride membrane , hybridized with primary antibody , reacted with horseradish peroxidase-conjugated secondary antibody , and visualized using chemiluminescent substrate ( Thermo Scientific ) . Cells were trypsinized , washed with PBS , and fixed in 70% ethanol . Fixed cells were double stained with propidium iodide ( PI ) for DNA and with anti-pUL44 antibody to identify infected cells , and analyzed by flow cytometry . Cells were gated for pUL44 and PI staining , and percentages of cells in each cell cycle compartment were calculated using FlowJo software . Total RNA was extracted with TRIzol ( Invitrogen ) , treated with TURBO DNA-free ( Ambion ) to remove DNA contaminants , and reverse transcribed with random hexamer primers using the High Capacity cDNA RT Kit ( Applied Biosystems ) . The cDNA was quantified by qPCR using SYBR green SYBR Advantage qPCR Premix ( Clontech ) with primers for the cellular genes Cyclin A or GAPDH ( glyceraldehyde-3-phosphate dehydrogenase ) ( Table 1 ) . cDNA from six arbitrary samples were mixed together and serially diluted to generate a standard curve used to calculate the relative amount of specific RNA . | Cyclins are evolutionarily conserved proteins that associate with cyclin-dependent kinases ( CDKs ) to regulate phosphorylation of multiple substrates to promote cell-cycle progression . Many viruses manipulate the cell cycle in order to create an environment suitable for replication; however , only few examples exist where viruses modulate cyclin activity . Here , we identified a cyclin-binding domain within the human cytomegalovirus ( HCMV ) protein pUL21a that confers its ability to interact with cyclin A and target it for proteasome degradation . Cyclin A promotes cellular DNA replication , which consumes important enzymes and metabolites needed for viral replication , making it important for large viruses like HCMV to block this protein's activity . In accord , the ability of pUL21a to degrade cyclin A was necessary for the virus to block cellular DNA replication and promote viral replication . Importantly , ablating cyclin A expression restored replication to a virus lacking pUL21a , demonstrating that cyclin A has the intrinsic ability to restrict viral replication , but is specifically countered by pUL21a . Together with our previous work showing that pUL21a also regulates the anaphase-promoting complex , another master cell cycle regulator , our studies have now revealed that HCMV has elegantly evolved dual functions within one protein targeting the cell cycle machinery for viral replication . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| []
| 2013 | Cyclin A Degradation by Primate Cytomegalovirus Protein pUL21a Counters Its Innate Restriction of Virus Replication |
A defect in the PKA1 gene encoding the catalytic subunit of cyclic adenosine 5′-monophosphate ( cAMP ) –dependent protein kinase A ( PKA ) is known to reduce capsule size and attenuate virulence in the fungal pathogen Cryptococcus neoformans . Conversely , loss of the PKA regulatory subunit encoded by pkr1 results in overproduction of capsule and hypervirulence . We compared the transcriptomes between the pka1 and pkr1 mutants and a wild-type strain , and found that PKA influences transcript levels for genes involved in cell wall synthesis , transport functions such as iron uptake , the tricarboxylic acid cycle , and glycolysis . Among the myriad of transcriptional changes in the mutants , we also identified differential expression of ribosomal protein genes , genes encoding stress and chaperone functions , and genes for secretory pathway components and phospholipid synthesis . The transcriptional influence of PKA on these functions was reminiscent of the linkage between transcription , endoplasmic reticulum stress , and the unfolded protein response in Saccharomyces cerevisiae . Functional analyses confirmed that the PKA mutants have a differential response to temperature stress , caffeine , and lithium , and that secretion inhibitors block capsule production . Importantly , we also found that lithium treatment limits capsule size , thus reinforcing potential connections between this virulence trait and inositol and phospholipid metabolism . In addition , deletion of a PKA-regulated gene , OVA1 , revealed an epistatic relationship with pka1 in the control of capsule size and melanin formation . OVA1 encodes a putative phosphatidylethanolamine-binding protein that appears to negatively influence capsule production and melanin accumulation . Overall , these findings support a role for PKA in regulating the delivery of virulence factors such as the capsular polysaccharide to the cell surface and serve to highlight the importance of secretion and phospholipid metabolism as potential targets for anti-cryptococcal therapy .
Cryptococcus neoformans is a basidiomycete fungal pathogen that infects both immunocompromised and immunocompetent individuals to cause meningioencephalitis [1] . A variety of virulence factors have been characterized , including the formation of a polysaccharide capsule , the production of the pigment melanin in the cell wall , the ability to grow at 37 °C , and the secretion of phospholipase , urease , and other extracellular components [2 , 3] . A common theme is that many of the virulence factors require transport to the plasma membrane or cell wall , or secretion to the extracellular environment . This is particularly true for the capsule polysaccharide that is considered to be the major virulence factor of the fungus [1 , 3 , 4] . The capsule is known to have a variety of immunomodulatory effects , and acapsular mutants are attenuated for virulence in animal models [3–7] . Many factors , such as iron starvation , serum , carbon dioxide levels , and pH and glucose levels , influence the size of the capsule in C . neoformans [4 , 8] . Protein trafficking is also important to localize the enzyme laccase to the cell wall for the polymerization of diphenol substrates to produce melanin [9] . Melanization in C . neoformans contributes to survival within alveolar macrophages , resistance to oxidative stress , and extra-pulmonary dissemination; melanin may also protect the fungus from environmental predators such as amoebae or from UV irradiation [10–15] . The cyclic adenosine 5′-monophosphate ( cAMP ) /protein kinase A ( PKA ) signaling pathway regulates capsule size , mating , melanin formation , and virulence in C . neoformans [16–19] . Several components of the pathway have been characterized , including the genes encoding a Gα protein ( Gpa1 ) , adenylyl cyclase ( Cac1 ) , a candidate G-protein–coupled receptor ( Gpr4 ) , phosphodiesterase ( Pde2 ) , and the catalytic ( Pka1 , Pka2 ) and regulatory ( Pkr1 ) subunits of PKA . Expression of GPA1 is induced by nitrogen limitation , and gpa1 mutants show attenuated capsule production , reduced melanin formation , sterility , and lower virulence [16] . Adenylyl cyclase mutants display the same phenotypes as gpa1 mutants [19] . Recently , Bahn et al . [20] identified Aca1 , an adenylyl cyclase–associated protein , which functions in parallel with Gpa1 to control Cac1 . The cAMP pathway is activated in part through Gpr4; interestingly , this receptor responds to amino acids and influences capsule but not melanin formation [21] . Xue et al . [21] speculated that a separate G-protein–coupled receptor and a hexose transporter could function upstream of the cAMP pathway to mediate the response to glucose in the context of melanin synthesis . The genes encoding the catalytic ( PKA1 , PKA2 ) and regulatory subunits ( PKR1 ) in C . neoformans have been disrupted in strains of both the A and D serotypes [17 , 18 , 20] . In a serotype A strain , the pka1 mutant is sterile , unable to produce melanin or capsule , and is avirulent; these phenotypes resemble those of the gpa1 and cac1 mutants [16 , 19] . Mutants defective in PKR1 may constitutively express PKA activity , independent of upstream signals . Disruption of PKR1 suppresses the capsule and melanin defects of the gpa1 mutant , causes cells to display an enlarged capsule phenotype , and results in hypervirulence [17] . In a serotype D strain , the Pka1 catalytic subunit is not required for mating , haploid fruiting , production of capsule or melanin , and virulence [17 , 18] . Instead , a second PKA catalytic subunit ( Pka2 ) , which is present in both serotype A and D strains , regulates mating , haploid fruiting , and virulence factor formation . Pka2 has no apparent role in mating and melanin production in the serotype A strain , indicating that differences exist for the roles of Pka1 and Pka2 in strains of the different serotypes [18 , 20] . Although components of the cAMP/PKA pathway have been identified , downstream targets have yet to be characterized in detail . A recent microarray experiment identified genes whose expression is dependent on Gpa1 and revealed that cAMP regulates multiple genes at the transcriptional level to control capsule and melanin production [22] . Importantly , the LAC2 gene was identified by this approach , and this gene encodes a second laccase that is adjacent in the genome to the more thoroughly characterized laccase gene , LAC1 . Like LAC1 , LAC2 transcript levels are induced by response to glucose deprivation [22] . The influence of the cAMP/PKA pathway on transcription has been investigated in several fungi , including Saccharomyces cerevisiae , Candida albicans , and Ustilago maydis [23–26] . Microarray analyses with mutants defective in each of the three catalytic subunits of PKA ( Tpk1–3 ) in S . cerevisiae revealed that Tpk2 negatively regulates genes for iron uptake and positively regulates genes for trehalose degradation and water homeostasis [26] . Tpk1 influenced the expression of genes for branched chain amino acid biosynthesis . In a related approach , Jones et al . [25] used arrays to examine the influence of constitutive expression of the cAMP/PKA pathway due to loss of the PDE2 gene encoding a cAMP phosphodiesterase . These results linked cAMP signaling to ribosome biogenesis and the response to stress including a connection with the unfolded protein response ( UPR ) pathway . Additionally , the expression of a number of genes encoding cell wall functions was altered in the mutant . In C . albicans , transcriptional profiling of a mutant defective in adenylyl cyclase linked cAMP signaling to ribosome biogenesis , metabolism , cell wall functions , the dimorphic transition , and the response to stress [23] . The influence on cell wall functions and the dimorphic transition was further confirmed through the analysis of mutants defective in pathway components . For example , deletion of the PDE2 gene encoding a high affinity cAMP phosphodiesterase results in higher sensitivity to heat shock and agents that challenge cell wall integrity , perturbation of hyphal and pseudohyphal development , and changes in sensitivity to antifungal drugs [27 , 28] . Transcriptional analysis by serial analysis gene expression ( SAGE ) of U . maydis mutants defective in the catalytic or regulatory subunits of PKA also revealed connections with ribosome biogenesis , morphogenesis , and metabolism . In addition , this analysis revealed PKA regulation of phosphate acquisition and a role for phosphate sensing in morphogenesis [24] . In this report , we used SAGE to examine the transcriptomes of mutants defective in the catalytic ( pka1 ) or the regulatory ( pkr1 ) subunits of PKA and found that links between PKA , ribosome biogenesis , stress response , and metabolic functions are conserved in C . neoformans . However , we also observed PKA regulation of components of the secretory pathway . Coupled with the observed changes in transcript levels for ribosomal proteins and stress chaperones , these SAGE results suggested that PKA plays a role in remodeling secretion to facilitate cell surface expression of virulence factors . These observations are consistent with recent reports that a defect in a secretory component Arf1 reduces capsule size [29] and that exocytosis and specialized vesicles mediate the secretion of capsule polysaccharide [30 , 31] . In addition , we found that PKA activity had specific effects on genes for capsule production , indicating that the kinase may act both transcriptionally and post-transcriptionally to influence virulence . Functional analyses confirmed that PKA regulates the response to temperature stress and that secretion inhibitors block capsule production . Finally , deletion of a PKA-regulated gene , OVA1 , revealed an epistatic relationship with pka1 in the control of capsule size and melanin formation . This gene encodes a putative phosphatidylethanolamine-binding protein ( PEBP ) that may link phospholipid metabolism , secretion , and cAMP signaling in C . neoformans .
Given the phenotypes of mutants defective in the catalytic and regulatory subunits of PKA [17 , 18] , we hypothesized that the cAMP/PKA pathway might regulate the expression of known ( i . e . , CAP and CAS genes [4] ) and novel genes for capsule formation and other virulence factors at the transcriptional level . To test this idea , we generated SAGE libraries for the wild-type ( WT ) strain H99 ( 71 , 067 tags ) and for the pka1 and pkr1 mutants ( 68 , 350 and 49 , 224 tags , respectively ) using cells grown in low iron medium ( LIM ) to induce capsule expression ( Table S1 ) . The SAGE profiles were compared to identify differential transcript levels ( p ≤ 0 . 01 ) : 599 tags were found to be differentially expressed when the WT library was compared with the pka1 library , 285 were differential between the WT and pkr1 libraries , and 419 were differential between pka1 and pkr1 ( Table S2 ) . Overall , the percentages of differential tags between the libraries ranged from 1 . 31% ( pkr1 versus WT ) to 2 . 64% ( pka1 versus WT ) , indicating substantial changes in the transcript profiles in the mutants . The 100 most abundant tags for each library are listed in Table S3 , and a global comparison of shared unique tag sequences among the three libraries is presented in Figure S1 . We annotated all of the differentially expressed tags with regard to the corresponding genes and then sorted the genes into functional categories . This analysis revealed that genes in several categories were affected by defects in PKA ( Tables 1–4; Tables S3–S5 ) . These genes encode known virulence proteins , ribosomal proteins and other components of the translation machinery , a large number of heat shock and stress proteins , protein trafficking components , cytoskeleton proteins , transporters , cell surface and extracellular proteins , and a variety of metabolic functions for phospholipid synthesis , the tricarboxylic acid cycle , and glycolysis . Overall , these observations reflect the conserved role of the cAMP pathway in coupling environmental sensing ( e . g . , of nutrient levels ) with metabolism and growth . We initially focused on categories containing known genes for capsule and melanin formation ( CAP and CAS genes ) or genes with known associations with virulence ( e . g . , response to oxidative stress ) . We found that tags for the capsule-related genes CAS35 and CAP10 [4] had reduced levels in the pka1 and pkr1 mutants compared with tags for the WT strain in the low iron condition ( Table 1 ) . Interestingly , a tag for the USX1 gene encoding UDP-xylose synthase for capsule xylosylation was up-regulated in both mutants [32] . However , other capsule genes such as CAP59 , CAP60 , and CAP64 ( and other CAS genes ) did not show significant expression in all libraries and thus could not be assessed for differential transcript levels . Although we did not detect tags for the LAC1 and LAC2 genes , we did find a tag for a putative multi-copper oxidase/ferro-O2-oxidoreductase gene ( acidic laccase ) at lower levels in both mutants , and we identified a putative copper transporter with a lower tag level in the pka1 mutant ( Table 3 ) . Previously , Pukkila-Worley et al . [22] found that several genes for capsule production and two laccase genes , LAC1 and LAC2 , were regulated by the cAMP pathway component Gpa1 . Laccase is a copper-dependent enzyme and there is evidence for a role for copper transport in influencing laccase activity [15 , 33] . Several other genes implicated in virulence , including genes for functions involved in oxidative , nitrosative , and temperature stress , showed differential tag levels in the PKA mutants . These included the SOD1 ( Cu/Zn superoxide dismutase ) gene that had reduced tag levels in both mutants ( Table 1 ) . The antioxidant function of Sod1 is critical for the virulence of the fungus , and deletion of SOD1 affects the expression of a number of virulence factors , including laccase , urease , and phospholipase [34–37] . Notably , a tag for the gene encoding a superoxide dismutase copper chaperone also showed a higher level in the mutants . Enzymes that counteract oxidative and nitrosative stress contribute to virulence in C . neoformans . In this context , we found that the tag for the FHB1 gene encoding flavohemoprotein [38] had a higher abundance in the pkr1 mutant library relative to the WT library . In addition , the TSA1 gene encoding thiol peroxidase plays a role in resistance to oxidative and nitrosative stress and is also necessary for virulence ( Table 1 ) [39] . Among genes known to be important for high-temperature growth , a tag for the TPS1 gene for trehalose-6-phosphate synthase was more abundant in the pkr1 and pka1 mutants . Tps1 is involved in the synthesis of the stress protectant sugar , trehalose , and contributes to the ability of the fungus to grow at 37 °C [2 , 40] . Amino acid metabolism genes such as ILV2 ( acetolactate synthase ) and the SPE3/LYS9 chimeric gene ( chimeric spermidine synthase/saccharopine dehydrogenase ) are also required for growth at elevated temperature and for full virulence [41 , 42] . Tags for both ILV2 and SPE3/LYS9 were more abundant in the pka1 mutant libraries . Cyclophilin A is also required for growth at host temperature [43] , and a tag for the corresponding gene had a higher abundance in the pka1 mutant library . Cyclophilins have peptidyl-prolyl isomerase activity and catalyze protein folding [44] . Finally , a tag for TOP1 ( topoisomerase I ) was elevated in the pka1 and pkr1 mutant libraries ( Table 1 ) . TOP1 is essential in C . neoformans and its regulation under stress conditions may have an impact during initiation of infection [45] . Overall , these results indicate that PKA influences transcript levels for several functions implicated in the response to stress and in virulence . However , the patterns of regulation , with higher or lower transcript levels in one or both mutants , indicate a complex influence of PKA on virulence-related functions that likely involves direct and indirect control mechanisms . For example , 377 tags had higher levels and 507 had lower levels in both mutant libraries relative to the WT library ( Table S2 ) . Similar complex patterns of regulation were also observed in the SAGE analysis of PKA mutants defective in catalytic or regulatory subunits in U . maydis [24] . Our analysis of the SAGE data revealed that ∼20 tags representing stress response genes were elevated in the pka1 mutant compared with those in the WT strain ( Table 2 ) . Some of these genes were also elevated in the pkr1 mutant relative to WT , and six were down-regulated relative to the WT library . These tags represent genes with sequence similarity to a number of heat shock proteins ( HSPs ) , including Hsp10 , Hsp12 , Hsp60 , Hsp70 , Hsp90 , and Sks2 . Coupled with the genes for other stress-responsive proteins , such as glutathione peroxidase , transaldolase , alpha-trehalose-phosphate synthase , thioredoxin-dependent peroxide reductase , and cyclophilin A , these results indicate a conserved connection between PKA and stress in C . neoformans . Many of these proteins are important for secretion as well as resistance to oxidative and nitrosative stress , and several have been linked to virulence in C . neoformans or other pathogens [36–39 , 43 , 46–48] . For example , glutathione peroxidases are important for defense against oxidative killing in bacterial pathogens such as Streptococcus pyogenes [49] . HSPs are produced in response to heat stress [50] , and the SAGE results predicted that the pka1 mutant would be more resistant to temperature and oxidative stress compared with the WT and pkr1 strains because of the higher levels of tags associated with the HSPs . We confirmed this prediction by showing that the pka1 mutant was more resistant to a 50 °C heat shock than the WT H99 strain and the pkr1 mutant ( Figure 1 ) . The difference between the pka1 mutant and the WT strain was particularly evident in the plate assay at 10 min of heat shock ( Figure 1A ) and at 30 min for the quantitative assay ( Figure 1B , p < 0 . 01 ) . The pkr1 mutant was notably more sensitive to heat shock than either of the other strains ( Figure 1 ) . It is interesting to note that the higher transcript levels for genes for heat shock and oxidative stress proteins in the pka1 mutant must not make a sufficient or appropriate contribution to virulence given the attenuation observed for this mutant [17] . Previously , Jones et al . [25] found that a mutation in PDE2 , encoding a cAMP phosphodiesterase , influenced the expression of components of the UPR pathway in S . cerevisiae . The data for C . neoformans suggest that PKA activity influences ribosome biogenesis and the expression of stress-related proteins such as endoplasmic reticulum ( ER ) chaperone components that may have a role in protein folding . We therefore examined the functional categories of genes from the SAGE data for those known to be positively and negatively regulated by ER stress and UPR in yeast [51 , 52] . As listed in Tables 2 and 3 , the tags for a number of predicted functions related to ER stress were influenced by the pka1 and/or pkr1 mutations , including myo-inositol phosphate synthase , protein disulfide isomerase , cyclophilin , and Hsp70 family member chaperones . Additionally , tags for genes in other functional categories related to the UPR were also influenced by the PKA mutations . These included genes for ribosomal proteins ( Table S4 ) and amino acid biosynthetic enzymes ( Table 1; Table S5 ) . In mammalian cells , a hallmark of ER stress is a reduction in protein synthesis to reduce ER loading , and our SAGE data revealed that tags matching genes for 48 ribosomal protein genes were elevated in the pkr1 library and/or reduced in the pka1 library ( Table S4 ) . A similar influence of cAMP signaling on ribosomal protein expression has been previously described in other fungi [23–25 , 27 , 53] . These observations raise the possibility that one component of the differential influence of the pka1 and pkr1 mutations on capsule size might be due to an influence on the ER and the trafficking of proteins and capsule components . In this light , we noted that the SAGE data contained a group of tags elevated in the pkr1 library and/or reduced in the pka1 library that identified genes for predicted proteins connected with protein or vesicle trafficking machinery ( Table 3 ) . These functions included a putative membrane protein required for vesicular transport ( Bet1 ) , a syntaxin , an ADP-ribosylation factor ( ARF ) GTPase activator , a protein–ER retention–related protein , a subunit of the protein transport protein Sec61 , a protein–vacuole targeting-related protein , a phosphomannomutase , and an exocyst complex subunit ( rsec15 ) . Two other tags identified a gene for a member of the Rab superfamily of Ras-related proteins ( Ypt3 ) and a gene for an ER–Golgi transport–related protein; these were elevated in the pka1 library relative to the pkr1 and/or WT libraries . In yeast and animal cells , trafficking of cargo molecules through the secretory pathway relies on packaging and delivery of membrane vesicles [54] . Bet1 and syntaxin are SNAREs ( soluble N-ethylmaleimide-sensitive factor attachment protein receptors ) that mediate vesicle fusion with target membranes in S . cerevisiae . The ARF GTPase activator protein stimulates the nucleotide exchange and hydrolysis reactions of ARF GTPases and therefore regulates Arf-mediated vesicle transport [55 , 56] . Disruption of the C . neoformans ARF1 gene encoding an ADP ribosylating factor has recently been shown to reduce capsule size [29] . Capsule elaboration must involve the trafficking of a large amount of polysaccharide material along with the proteins necessary for assembly to the cell surface [30 , 31] . Based on the SAGE data , we hypothesized that PKA could influence capsule formation by regulating secretion . We therefore examined the effects of known vesicle trafficking inhibitors that target functions identified by the SAGE data on cell growth , morphology , and capsule formation . These inhibitors included brefeldin A ( BFA ) , nocodazole ( NOC ) , monensin , and N-ethylmaleimide ( NEM ) . BFA is known to arrest the anterograde transport of proteins between the ER and Golgi apparatus by interfering with the action of ARF . As mentioned , the SAGE data revealed that a gene for a putative ARF–GTPase-activating protein ( GAP ) was regulated by PKA . NOC inhibits vesicle trafficking by interfering with microtubule formation , and we were prompted to test this compound because the transcript levels for α and β tubulin were elevated in the pka1 mutant ( Table S5 ) . Monensin is a Na+/H+ ionophore that blocks intracellular transport in both the trans-Golgi and post-Golgi compartments . NEM is a cysteine alkylating agent that interferes with disulfide bond formation . The inhibitors did not block growth when tested with cells on yeast extract peptone dextrose ( YPD ) plates for 3 d ( for NOC , cells were grown in either liquid YPD or low iron medium ) , although the cells grew slower in the presence of NEM or BFA ( Figure 2A; unpublished data ) . The WT and mutant strains also grew in the presence of the iron chelator bathophenanthroline disulfonic acid ( BPS ) with or without the addition of monensin ( Figure 2A ) . These conditions were tested because LIM was employed to assess capsule formation . In this regard , capsule size was significantly reduced ( p < 0 . 001 ) when the cells were grown at different concentrations of the inhibitors in LIM at 30 °C ( Figure 2B–2D ) . This result was found both for WT , and for the pkr1 mutant that otherwise displays an enlarged capsule . Additionally , an abnormal morphology was observed for cells of both strains in LIM with 10 or 25 μg/ml of NOC in that mother and daughter cells failed to separate ( Figure 2B ) . The influence on capsule size was more pronounced at higher concentrations of inhibitors and with longer treatment for both strains ( 48 h ) . The SAGE data and the inhibition assays together support a model in which connections between PKA and the secretory pathway explain , at least in part , why the loss of PKA activity results in a small capsule [17] . A key aspect of trafficking is the biosynthesis of phospholipids to support membrane formation for the ER , Golgi , and vesicular network [57] . Inositol is an essential component of phosphospholipids as well as phosphoinositides that act as second messengers in signaling cascades and that regulate membrane trafficking [58–60] . Additionally , INO1 encoding inositol 1-phosphate synthase is a well-characterized target of both UPR and PKA regulation in S . cerevisiae [61 , 62] . Genes related to inositol metabolism were identified by SAGE including the C . neoformans ortholog of INO1 encoding myo-inositol 1-phosphate synthase; the tag for this gene had a reduced level in the pkr1 mutant ( Table 3 ) . A tag for a gene encoding a putative inositol-1- ( or 4 ) -monophosphatase was found to have lower levels in the pka1 and pkr1 mutants relative to WT . This enzyme is important for converting glucose 6-phosphate to myo-inositol phosphate . A tag for a gene encoding phosphatidylglycerol/phosphatidylinositol transfer protein ( PITP ) was elevated in both the pkr1 and WT libraries relative to the pka1 library . PITP was named due to its ability to transport phosphatidylinositol between membrane compartments , and recent data indicate that PITP plays an important role in the coupling of PIP2 ( phosphatidylinositol-4 , 5-bisphosphate , a substrate for phospholipase C ) to signaling and membrane trafficking in mammalian cells [63 , 64] . In S . cerevisiae , PITP is required for cell viability [63 , 64] , and in Yarrowia lipolytica , PITP is required for differentiation from the yeast to the mycelial growth form [63] . The SAGE data also identified a tag for an ortholog of the S . cerevisiae ITR1 gene that encodes a myo-inositol transporter protein , the major permease for inositol uptake [65] . Taken together , the data indicate that PKA regulates genes whose products influence the level of free inositol in the cell , and inositol is known to play a major role in the regulation of phospholipid biosynthesis [66 , 67] . To functionally examine the role of inositol metabolism ( and phospholipid biosynthesis indirectly ) in the control of capsule production , we tested the influence of lithium on growth and capsule size . Lithium is an inhibitor of inositol monophosphatase and may also influence inositol uptake as well as other cellular processes [68 , 69] . We found that the pka1 mutant was more sensitive to 75 mM and 150 mM lithium chloride ( LiCl ) , especially at 37 °C , compared with the pkr1 mutant and the WT strain ( Figure 3A ) . The sensitivity of the pka1 mutant is consistent with reduced transcript levels for genes encoding a myo-inositol transporter and the inositol monophosphatase in the mutant relative to the WT strain . Changes in transcript levels for these genes were also observed in the pkr1 mutant ( Table 3 ) , and we noted slightly reduced growth for this mutant in 150 mM LiCl at 37 °C . Given the capsule defect in the pka1 mutant [17] and the influence of hypertonic salt solutions on capsule size [70] , we reasoned that LiCl might also influence capsule formation in WT cells and the pkr1 mutant , and this was the case . As shown in Figure 3B and 3C , treatment with a range of relatively low LiCl concentrations ( 1 mM to 75 mM ) reduced capsule size in cells grown in LIM in a dose-dependent manner . This result supports the idea that inositol and perhaps phospholipid metabolism are important for trafficking capsule material , although it is also possible that lithium influences other processes in C . neoformans . Growth in the presence of glycerol is also known to influence phospholipid metabolism , and glycerol can act as a chemical chaperone to influence protein trafficking indirectly by mediating protein folding and transport [71 , 72] . We therefore examined the influence of glycerol in the WT and mutant cells and found inhibition of capsule formation ( Figure 3D and 3E ) . This result further supports the hypothesis that phospholipid biosynthesis and vesicle trafficking are important for capsule formation . Several differential tags identified genes encoding proteins targeted to intracellular organelles , the plasma membrane , or the cell surface . For example , 13 of these tags matched genes encoding membrane-associated transporters ( Table 3 ) . These included transporters related to carbohydrate import/export such as two monosaccharide transporters ( hexose transport-related protein , glucose transporter ) and the myo-inositol transporter that are elevated in pkr1 library and/or reduced in pka1 library ( Table 3 ) . A tag for hexose transporter gene was also elevated in the pka1 library relative to the WT and pkr1 libraries . The transport and sensing of sugars may be a critical component of both the cAMP signaling pathway and the acquisition of substrate to support capsule formation . For example , genes for glucose transporters were expressed during cryptococcal experimental meningitis and interaction of the fungus with macrophages [73 , 74] , and a hexose transporter has been proposed to be upstream of the cAMP pathway in C . neoformans [21] . In addition , one putative hexose transporter has been functionally characterized and found to be involved in capsule formation , but not virulence , in a Caenorhabditis elegans killing test [75] . Other tags identified genes for a group of transporters related to phosphate uptake and assimilation . A transcript for a phosphate transporter was elevated in the pka1 library , and a tag for a gene encoding putative sodium:inorganic phosphate symporter was reduced in both the pka1 and pkr1 libraries compared with the WT library . These observations are similar to our description of an influence of PKA on the expression of phosphate acquisition and storage functions in U . maydis [24 , 76] . Additional tags in this group identified a gene for a urea transporter that was elevated in the WT library and a tag for a metal ion transport–related protein that was lower in the pka1 library . Transcripts encoding proteins involved in cell wall synthesis and having an extracellular location were also affected in the pka1 or pkr1 mutants ( Table 4 ) . In this category , eight tags were elevated and one tag was reduced in the pka1 mutant . Some of the corresponding genes encode a chitin deacetylase , an 88-kDa mannoprotein ( MP88 ) , and a polypeptide with similarity to the OV-16 antigen presursor that we designated Ova1 [77] . MP88 may be relevant to virulence because this mannoprotein is known to stimulate T-cell responses [77] . Other tags represented genes for putative enzymes involved in cell wall synthesis or remodeling; these enzymes included an α-1 , 3-glucan synthase ( Ags1 ) , a β-1 , 3-glucan synthase ( Fks1 ) , an endoglucanase , an α-1 , 4 glucan branching enzyme , and an endo-1 , 3 ( 4 ) -β glucanase . The influence of PKA on candidate genes for cell wall functions prompted an examination of the sensitivity of the strains to agents that challenge cell wall integrity such as SDS , Congo red , calcofluor white , and caffeine . Significant differences between the WT strain and the pka1 and pkr1 mutants were not found except for caffeine , where the pka1 mutant grew more slowly than the WT strain and the pkr1 mutant at 37 °C ( Figure 4 , and unpublished data ) . Caffeine affects many cellular processes , including cAMP phosphodiesterase activity , and has been used to detect cell integrity phenotypes in S . cerevisiae , especially in the context of the mitogen-activated protein kinase/cell wall integrity pathway [78] . The increased sensitivity of the pka1 mutant to caffeine supports the idea that cAMP signaling is involved in the maintenance of cell wall integrity in C . neoformans , although additional influences of caffeine cannot be ruled out . We also examined the sensitivity of strains to osmotic stress and did not observe differences on medium with 1 . 5 M NaCl , 1 . 5 M KCl , or 1 . 5 M sorbitol ( unpublished data ) . The discovery of a connection between secretion , PKA , and capsule formation in C . neoformans suggests that specific downstream targets of PKA may have regulatory roles that influence capsule size . As indicated earlier , we identified a gene , OVA1 , with an elevated tag count in the pka1 mutant , indicating that PKA has a negative influence on the transcription of the gene . OVA1 is also of particular interest because the tag was abundant in a SAGE library prepared with cells from the cerebral spinal fluid of infected rabbits [74] , and the gene encodes a predicted polypeptide with similarity to the conserved PEBP family present in many organisms such as mammals , fungi , worms , and bacteria ( Figure 5 ) [79–82] . Ova1 also shows similarity to the OV-16 antigen of the river blindness parasite Onchocerca volvulus and was identified as a mannoprotein in C . neoformans by Huang et al . [77] , indicating that the protein is secreted . In S . cerevisiae , the most similar protein ( a yeast PEBP ) , Tfs1 ( for “twenty-five suppressor” ) , was isolated as multicopy suppressor of the cdc25–1 mutant [82] . Cdc25 in yeast is one of the Ras guanine exchange factors ( GEFs ) that activates Ras to subsequently stimulate adenylyl cyclase . Tfs1 also inhibits Ira2 , a Ras GAP in yeast , thus making an additional connection with the cAMP pathway . Furthermore , Tfs1 is an inhibitor of carboxypeptidase Y , a well-characterized cargo protein for examining trafficking in S . cerevisiae . Overall , these observations suggest a conserved connection between putative PEPB proteins and cAMP signaling in fungi , and lead us to hypothesize that Ova1 functions in the secretory pathway to influence trafficking functions for capsule formation . We initially confirmed that OVA1 shows a higher transcript level in the pka1 mutant compared with the WT strain and the pkr1 mutant by RNA blot analysis ( Figure 6 ) and real-time PCR analysis ( Figure S2 ) . Given that the SAGE libraries were prepared with cells from LIM , we also examined the influence of iron on OVA1 transcript levels and found no effect ( Figure 6 ) . We subsequently generated deletion mutants for OVA1 as well as reconstituted strains in which OVA1 was reintroduced to complement the mutation . The ova1 deletion mutants did not show significant growth differences at 30 °C or 37 °C compared to WT , but did display a slightly enlarged capsule in different inducing media , including LIM ( Figure 7A and 7B ) . We included the ova1 mutant in our tests of the influence of LiCl and glycerol on capsule formation and found that the mutant was as sensitive as the WT strain and the pkr1 mutant ( Figure 3C and 3D ) . The OVA1 gene was also disrupted in the pka1 mutant background to examine epistasis with regard to capsule and melanin formation . Interestingly , the pka1 ova1 double mutant showed partial restoration of capsule formation compared to the pka1 mutant in both LIM and Dulbecco's modified Eagle's medium ( DMEM ) , suggesting that Ova1 functions downstream of PKA to influence capsule size ( Figure 7A and 7B ) . The double mutant additionally showed partial restoration of melanin production compared to the pka1 mutant on medium containing dopamine as a substrate ( Figure 7C ) . We also tested the pka1 ova1 mutant for sensitivity to lithium and found enhanced sensitivity compared to either of the single mutants ( Figure 7D ) . The WT or the pka1 phenotypes were restored upon reintroduction of the OVA1 gene into the ova1 mutant or the pka1 ova1 double mutant , respectively . Given the putative phosphatidylethanolamine-binding domain , the lithium sensitivity of the double mutant , and the influence of PKA on transcript levels for genes in inositol metabolism , we propose that Ova1 plays a negative role in phospholipid-related trafficking functions needed for capsule production and potentially influences laccase transport . Of course , it is also possible that Ova1 functions in other processes to influence capsule size and melanin production in the background of the pka1 mutation . Finally , we performed a preliminary test of the role of Ova1 in virulence because of our observations that ova1 mutants had a slightly enlarged capsule . We inoculated the mutant , the WT , and the complemented strain into mice , and we found no defect in the ability of the mutant to cause a lethal infection ( unpublished data ) .
The cAMP pathway regulates a variety of processes in fungi , including nutrient sensing , growth , the response to stress , and morphogenesis [83 , 84] . For C . neoformans , Alspaugh et al . [16] and D'Souza et al . [17] showed that this pathway controls formation of the polysaccharide capsule that is the major virulence factor of the fungus . We therefore examined the influence of mutations in genes encoding a catalytic subunit or the regulatory subunit of PKA on the transcriptomes of cells from capsule-inducing medium to gain insight into the mechanisms underlying changes in capsule phenotype . We found that defects in PKA had effects on transcript levels for genes involved in virulence , ribosome biogenesis , the response to stress , vesicle ( protein ) trafficking , membrane transport , and cell wall biogenesis . Although some of these genes represent conserved targets of cAMP signaling in fungi , our analysis also revealed a novel relationship between cAMP signaling and the secretory pathway in C . neoformans with coordinated changes in ribosomal protein and heat shock gene expression . This discovery focused our attention on PKA-regulated secretion as a potential central component of virulence factor elaboration . In particular , transcriptional changes in the pka1 and pkr1 mutants indicated an influence at several stages in the secretory pathway , including translocation ( Sec61 and Hsp70/Kar2 ) , maturation in the ER ( Hsp70/Kar2 , protein disulfide isomerase ) , vesicle formation and fusion ( Bet1 , syntaxin ) , Golgi transport ( α-1 , 6-mannosyltransferase , phosphatidylglycerol/phophatidylinositol trans-fer protein ) , and vesicle delivery to the plasma membrane ( e . g . , Ypt3 ) . Support for a key role of PKA in the regulation of secretion also came from observed transcriptional changes for genes encoding chaperones and enzymes for phospholipid metabolism . Treatment with inhibitors of protein secretion and chemicals that influence phospholipid metabolism ( lithium and glycerol ) confirmed a role in capsule elaboration . These results may partially explain observations in the literature such as the finding of Jacobson et al . [70] that a high salt concentration suppresses capsule size; salt stress is known to influence phospholipid metabolism and the accumulation of compatible solutes such as glycerol in yeast [85] . The influence of defects in PKA on the transcriptome in C . neoformans suggests a model in which PKA regulates the expression of secretory pathway components to control the elaboration of virulence factors at the cell surface . In particular , we believe that changes in the regulation of PKA activity , as would likely be found in the pkr1 mutant , could mediate a remodeling of the secretory pathway to accommodate the delivery of large amounts of polysaccharide material . Additionally , PKA could directly influence the activity or localization of transcription factors that control the transcript levels of some of the genes detected by SAGE , for example , in response to nutrient availability ( e . g . , glucose , nitrogen , iron , phosphate ) . A paradigm for this scenario exists with the influence of PKA on the Opi1p repressor of inositol synthesis in yeast [86] . In C . neoformans , phosphorylation by PKA could also positively or negatively influence the activity of proteins in the secretory pathway , and this may influence a UPR-like response to indirectly regulate transcription factors leading to the observed transcriptional responses . In addition to transcriptional effects , a more direct influence is also possible because PKA has been shown to phosphorylate Sso exocytic t-SNAREs to inhibit SNARE assembly and vesicle fusion in S . cerevisiae [87 , 88] . Furthermore , PKA has been shown to phosphorylate secretory components such as cysteine string proteins ( CSPs ) ( DNAj co-chaperone ) , Snapin , Rim1 , Snap-25 , syntaphilin , and synapsin in higher eukaryotes [89] . The cAMP/PKA pathway also regulates aspects of autophagy , a process linked to the UPR in yeast [90] . In general , the UPR pathway in S . cerevisiae provides a useful paradigm for the connection between molecular events in the ER that regulate secretion and that signal to the nucleus to cause transcriptional changes [51 , 52] . The UPR results in a transcriptional influence on ∼400 genes: regulated functions include ER chaperones , phospholipid biosythesis , ER-associated protein degradation ( ERAD ) , cell wall components , and anti-oxidative stress proteins; hallmark target genes in yeast include INO1 , KAR2 , and PDI1 [51 , 52] . We observed transcriptional changes in similar functions in the PKA mutants for C . neoformans . For example , genes involved in the response to oxidative stress and heat shock , and genes for chaperones and ribosomal proteins represented some of the largest groups that were differentially transcribed in the mutants . We also noted differential transcripts for functions associated with ERAD such as genes for putative ubiquitin ligases that showed elevated transcripts in the pka1 library ( G . Hu , unpublished data ) . The variety of potential levels of regulation will make it challenging to establish the mechanisms for specific targets of PKA control in C . neoformans . There is a growing body of evidence to link capsule synthesis and secretion in C . neoformans . Capsule biosynthesis may take place in the ER and the Golgi as suggested by subcellular studies that identified “wall vesicles” as potential structures involved in capsule biosynthesis and/or secretion [91–97] . Alternatively , capsule synthesis could occur at the cell membrane [91–97] . Walton et al . [29] showed that disruption of the ARF1 gene , which encodes the ARF GTPase involved in vesicle transport , reduced capsule size . In addition , Yoneda and Doering [30] found that mutation of a SEC4/RAB8 homolog resulted in the accumulation of vesicles containing material that stained with anti-capsular antibody . These authors concluded that capsule material is synthesized internally and secreted by exocytosis . More recently , Rodrigues et al . [31] showed that C . neoformans cells in culture and from infected macrophages produce extracellular vesicles that contain capsule polysaccharide . These vesicles are believed to provide a mechanism for moving high molecular weight capsule material across the cell wall . Two capsule-associated gene products , Cap10 and Cap60 , are localized to intracellular vesicles and to a compartment adjacent to the nuclear envelope , respectively [5–7] . Another capsule-associated gene , CAP59 , may also participate in polysaccharide export [94] . Connections between membrane trafficking and virulence in C . neoformans have also been established by Erickson et al . [98] . They characterized the VPH1 gene encoding a vacuolar ( H+ ) -ATPase and found that disruption of this gene results in defects in capsule formation , laccase production , urease production , and growth at 37 °C . A role in vesicular acidification was proposed to contribute to trafficking of capsule polysaccharide and laccase , with additional potential roles in signal sequence processing or glycosylation . Secretion would also potentially influence melanin production by influencing the localization of laccase to the cell wall . Based on the SAGE results , we searched the set of PKA-regulated genes for those that might play a role in secretion and that might have a connection with cAMP signaling in other organisms . One gene , OVA1 , encoded a predicted extracellular mannoprotein [77] with similarity to PEBPs and to the Tfs1p protein of S . cerevisiae [82] . PEBPs play several roles in mammalian cells , including lipid binding , inhibition of serine proteases , and regulation of signaling components such as heterotrimeric G proteins , as well as other functions [99] . Lipid ( e . g . , oxysterol ) binding proteins are known to play a role in vesicle transport [100] , and PKA could potentially regulate vesicle transport through an influence on these proteins . In this context , interconnections with phospholipid metabolism are possible because , as mentioned above , the transcription factor Opi1p acts as a lipid sensor and is a target of PKA phosphorylation [101] . The similarity of OVA1 to TFS1 is interesting because of a shared connection with PKA . Specifically , Tfs1p was initially identified as a multicopy suppressor of the cdc25–1 mutation in S . cerevisiae [82] . More recent work has shown that Tfs1p is an inhibitor of the Ras-GAP encoded by IRA2 and carboxypeptidase Y ( CPY is a well-characterized protein that traffics through the yeast secretory pathway ) [102 , 103] . Given that CDC25 encodes the Ras-GEF and IRA2 encodes the Ras-GAP , and that Ras functions in S . cerevisiae to stimulate cAMP signaling , Tfs1p appears to be an activator of the RAS/cAMP/PKA pathway that establishes links to lipid binding and potential secretion functions . TFS1 is also overexpressed in response to oxidative stress , diauxic shift , and heat shock , and contains a stress response element ( STRE ) , indicating transcriptional control by Msn2/Msn4 [102] . Furthermore , loss of Tfs1 suppresses growth inhibition caused by caffeine [102] . The functional similarities between Tfs1 and Ova1 and the shared PEBP domain suggested that Ova1 might play a role in linking PKA and secretion in C . neoformans . The SAGE data support this idea because the OVA1 transcript was found to be elevated in the pka1 mutant along with the heat shock/stress gene transcripts suggesting that OVA1 might also be stress-responsive like TFS1 . Deletion of OVA1 partially restored capsule and melanin formation in a pka1 mutant indicating Ova1 functions downstream of PKA and has a negative influence on capsule size and melanin accumulation . These observations suggest the hypothesis that Ova1 plays a regulatory role in the trafficking of protein and polysaccharide to the cell surface , perhaps as a component of secretory vesicles . Ova1 is predicted to carry a glycosylphosphatidylinositol ( GPI ) anchor that may serve to attach the protein to either the cell membrane or β-1 , 6-glucans in the cell wall . Although we did not observe significant defects in cell wall integrity in the ova1 mutant ( unpublished data ) , it is known that cAMP signaling is required in S . cerevisiae and C . albicans for maintenance of cell wall integrity [23 , 28 , 104] . There may be issues of redundancy to consider because another gene ( OVA2 ) that is predicted to encode a PEBP is present in the C . neoformans genome . In summary , a common theme in pathogenesis is the elaboration of extracellular and cell surface–associated virulence factors by pathogens . We propose that the cAMP pathway is critical for coordinating nutrient sensing with secretion in C . neoformans , particularly during infection . The SAGE analysis provides target genes that will be valuable for investigating the expression of secretory system components and virulence factors in the context of cryptococcal growth in mammalian hosts . Importantly , many of the genes that we have identified for secretion , heat shock , and transport showed abundant messages in the same strain of C . neoformans isolated from the cerebral spinal fluid of a rabbit model of cryptococcal meningitis [74] . These observations provide confidence that the in vitro conditions used for SAGE analysis of virulence factor regulation have relevance to infection . We should note that our SAGE analysis provides a view of the regulation of gene expression by PKA only in the serotype A background of C . neoformans . It is known that differences exist in PKA signaling between strains of the A and D serotypes [18] . Therefore , additional work is needed to explore whether the regulatory properties that we discovered are generally applicable to serotype D strains and other pathogenic Cryptococcus species . Finally , a more detailed understanding of functions that regulate capsule formation in C . neoformans may ultimately contribute to improved therapy . The capsule is an important therapeutic target because of its central role in virulence; in particular , accumulation of polysaccharide in the cerebral spinal fluid of patients is thought to be a contributing factor in the development of elevated intracranial pressure that results in neurological symptoms during cryptococcal meningitis [105 , 106] .
The C . neoformans var . grubii strain H99 ( WT ) and the derived mutants with defects in PKA1 , PKA2 , or PKR1 [17] were generously provided by J . Heitman ( Duke University ) . For SAGE library construction , cells were grown for 3 d at 30 °C on YPD ( 1% yeast extract , 2% Bacto peptone , 2% dextrose ) plates from frozen stocks . A single colony was used to inoculate 5 mL of LIM prepared as described previously [107] . These cultures were grown overnight at 37 °C , and the cells ( H99 , 3 . 0 × 106 CFU/mL; pka1 , 3 . 0 × 106 CFU/mL; pkr1 , 1 . 2 × 106 CFU/mL ) were washed with sterile distilled water and inoculated into 45 mL of LIM for subsequent growth for 6 h at 37 °C . This time point was chosen to be consistent with previous SAGE experiments that identified iron-responsive genes [107] . Cells ( H99 , 5 . 0 × 107 CFU/mL; pka1 , 2 . 0 × 107 CFU/mL; pkr1 , 3 . 0 × 107 CFU/mL ) were harvested by centrifugation and flash frozen in an ethanol dry-ice bath before lyophilization overnight at −80 °C . L-DOPA medium was prepared as described [17] . RNA was isolated from lyophilized cells by vortexing with glass beads ( 3 . 0 mm , acid-washed and RNase-free ) for 15 min in 15 ml of TRIZOL extraction buffer ( Invitrogen , http://www . invitrogen . com ) . The mixture was incubated for 15 min at room temperature , total RNA was isolated according to the manufacturer's instructions ( Invitrogen ) , and RNA quality was assessed by agarose gel electrophoresis . Total RNA was used directly for SAGE library construction as described by Velculescu et al . [108] using the I-SAGE kit ( Invitrogen ) . The tagging enzyme for cDNA digestion was NlaIII , and 29 PCR cycles were performed to amplify the ditags during library construction . Colonies were screened by PCR ( M13F and M13R primers ) to assess the average clone insert size and the percentage of recombinants . Clones from the libraries were sequenced by BigDye primer cycle sequencing on an ABI PRISM 3700 DNA analyzer ( AME Bioscience , http://www . amebioscience . com ) . Sequence chromatograms were processed using PHRED [109 , 110] , and vector sequence was detected using Cross_match [111] . Fourteen-base-pair tags were extracted from the vector-clipped sequence , and an overall quality score for each tag was derived based on the cumulative PHRED score . Duplicate ditags and linker sequences were removed as described previously [107 , 108] . Only tags with a predicted accuracy of ≥99% were used in this study , and statistical differences between tag abundance in different libraries were determined using the methods of Audic and Claverie [112] . An overview of the abundance classes for the three libraries is presented in Table S1 . The number of different tag sequences and the total numbers of tags present in each abundance class for the WT , pka1 , and pkr1 strains are indicated . For preliminary assignment of tags to genes , we used the EST database available for strain H99 at the University of Okalahoma's Advanced Center for Genome Technology ( http://www . genome . ou . edu/cneo . html ) . When an EST sequence could not be identified for a particular tag , we used the genomic sequence available for H99 at the Duke University Center for Genome Technology ( http://cneo . genetics . duke . edu/data/index . html ) and the Broad Institute ( http://www . broad . mit . edu/cgi-bin/annotation/fungi/cryptococcus_neoformans ) to identify contigs with unambiguous tag assignments . Of the 599 unique tag sequences found to have differential abundance between the WT , pka1 , and pkr1 libraries at a threshold p-value of less than 0 . 01 ( Table S1 ) , 25 were found to match two or more different locations in the genome sequence and were not included for further analysis . An additional 76 tags did not match any of sequences in either the genomic or EST databases for strain H99 . These tags may result from reverse transcription or sequencing errors , incomplete H99 genomic or EST sequence data , or tag overlap of an intron position . The remaining tags could be unambiguously assigned to candidate transcripts , and the corresponding EST or genomic sequence was used to search the nonredundant database at the National Center for Biotechnology Information ( NCBI ) using BLASTx ( Basic Local Alignment Search Tool ) . Each BLASTx result was inspected individually and recorded to prepare tables of tags and the corresponding predicted genes . In the case of genomic DNA sequences where introns are present , the expected values recorded were higher than would have been expected if the introns had been removed . The same was true for the results determined for both genomic sequences and ESTs compared to those that would have been expected if the sequences had been translated and the BLASTp algorithum were employed rather than a BLASTx algorithm . To examine the response of C . neoformans to stress , exponentially growing cultures were washed , resuspended in H2O , and adjusted to 106 cells/ml . The cell suspensions were the diluted 10-fold serially , spotted onto YPD medium supplemented with or without 1 . 5M KCl , 1 . 5M NaCl , 75mM , or 150mM LiCl , 0 . 5 mg/ml caffeine , 0 . 01% and 0 . 1% SDS , 0 . 5 or 1 mg/ml calcofluor white ( Fluorescent Brighter 28 ) , 1 . 5 M sorbitol , or 0 . 5 mg/ml Congo red . Plates were incubated for 3–4 d at 30 °C or 37 °C , and photographed . For heat shock treatment , early log phase cells grown at 30 °C were adjusted with YPD to 105 cells/ml , and incubated in a 50 °C water bath for 0 , 2 , 3 , 5 , 10 , 30 , or 60 min , and 4 μl of cells were spotted onto YPD plates . The plates were incubated at 30 °C and growth was monitored for 3 d . Quantitative analysis was performed by plating cell dilutions to determine colony forming units . An ova1::NEO disruption allele was constructed using the following primers and a modified overlap PCR procedure [113 , 114] . Briefly , the primers hug2–1/hug2–3 ( CACGTGCCAAGACTGAACAT/ AGCTCACATCCTCGCAGCAGACAAGGGAGGGTCAAGG ) and hug2–4/hug2–6 ( TAGTTTCTACATCTCTTCCTCACAACGGAAAGGACGAT/ TATTCGCGGCTATTTGGAAC ) were used with genomic DNA to obtain the left ( ∼1 kb ) and right ( ∼1 kb ) arms for the disruption construct . The selectable marker NEO was amplified using the primers hug2–2/hug2–5 ( CCTTGACCCTCCCTTGTCTGCTGCGAGGATGTGAGCT/ ATCGTCCTTTCCGTTGTGAGGAAGAGATGTAGAAACTA ) and the plasmid pJAF1 ( J . Heitman ) , which contains the neomycin antibiotic resistance marker cassette for C . neoformans . The ova1::NEO allele results in the deletion of the complete open reading frame of OVA1 ( ∼1 . 5 kb ) . The resulting 3 . 5-kb PCR product was used to transform strain H99 by biolistic transformation . Transformants were screened by colony PCR for the ova1::NEO allele using primers hug2-IntF/hug2-IntR ( negative screen ) ( GCTCACAACACCGACAAC/GGAGACTTGATCACTGCGA and hug2–9/hug-NEO ( CCAGCGATGCCATTTCCAT/AGCTCACATCCTCGCAGC ) ( positive screen ) . Primer hug2–9 was designed from the region upstream of OVA1 and hug-NEO was designed for the NEO gene . Transformants in which the WT allele was replaced were confirmed by Southern blot hybridization . Three mutants containing the allele designated ova1Δ were studied further . The OVA1 gene for complementation of the ova1Δ mutant was amplified by PCR using primers ova1-BamH1–5 ( CAGGGATCCAACCGTTCCATCAGGATGAC ) and ova1-BamH1–3 ( AGAGGATCCGCCAAGCGGTTCAATATAAGG ) , and H99 genomic DNA . The ∼3 . 35-kb product was digested with BamHI and cloned into the BamHI site of pCH233 , creating plasmid pHG100 . Strains ova1Δ-47 and pka1Δova1Δ–2 were transformed with pHG100 by biolistic transformation [107] , and reintroduction of OVA1 was confirmed by colony PCR and Southern blot analysis ( unpublished data ) . Total RNA was isolated as presented above and hybridization was performed as previously described [107] . The hybridization probe was prepared with a PCR-amplified DNA fragment of OVA1 using the specific primers hug2-IntF and hug2-IntR , and lableled with 32P using an Oligolabelling kit ( Amersham Biosciences , http://www . amershambiosciences . com ) . Scanned images were analyzed using an AlphaImager 3400 ( Alpha Innotech , http://www . alphainnotech . com ) . Real-time PCR analysis was conducted using primers targeted to 3′ ends of the transcripts; primers were designed using PrimerExpress ( Applied Biosystems , http://www . appliedbiosystems . com ) . Total RNA from the frozen cells was extracted as described above , DNA was removed by treatment with DNase I for 30 min at 25 °C , and cDNA was synthesized using random hexamers and Superscript transcriptase II ( Invitrogen Canada ) . The resulting cDNA was used for real-time PCR with Power SYBR Green PCR mix ( Applied Biosystems ) according to the manufacturer's recommendations . An Applied Biosystems 7500 Fast Real-Time PCR System was used to detect and quantify the PCR products using the following conditions: incubation at 95 °C for 10 min followed by 40 cycles of 95 °C for 15 sec , and 60 °C for 1 min . The cDNAs of the ACT1 and GPD genes were used to normalize the data ( Table S6; [22 , 73] ) . Dissociation analysis on all PCR reactions confirmed the amplification of a single product for each primer pair and the lack of primer dimerization ( Applied Biosystems ) . Primers used for each gene are listed in Table S6 . Relative gene expression was quantified using SDS software 1 . 3 . 1 ( Applied Biosystems ) . A single colony from solid YPD medium for each strain was cultured overnight at 30 °C in liquid YPD medium . Two inducing media ( LIM and agar-based DMEM ) [107 , 115] were used to examine capsule formation . For DMEM , 3 × 105 cells were spotted onto the plate , and incubated for 72 h at 30 °C . For LIM , an overnight culture in liquid YPD medium was harvested and diluted in low iron water , and 106 cells were added into 3 ml of liquid LIM for further incubation at 30 °C for 48 h . After incubation , the capsule was stained by India ink and examined by differential interference microscopy ( DIC ) . A single colony on solid YPD medium for each strain was cultured overnight at 30 °C in liquid YPD medium . Cells were collected and washed with H2O , and 106 cells were added into 3 ml of liquid LIM containing different concentrations of vesicle trafficking inhibitors ( as indicated ) and further cultured in a shaking incubator at 30 °C . Cells were examined at different time points ( 16 , 48 , and 72 h ) by DIC after staining with India ink . Vesicle trafficking inhibitors included BFA , NOC , monensin , and NEM . LIM was also generated by the addition of BPS to 75 mM . All chemicals were obtained from Sigma ( http://www . sigmaaldrich . com ) . Statistical analysis of capsule size was performed using Student's t-test .
The GenBank ( http://www . ncbi . nlm . nih . gov ) accession numbers for the PEBP proteins discussed in this paper are human ( P30086 ) ; mouse ( AF300422_1 ) ; Onchocerca volvulus ( P31729 ) ; Saccharomyces cerevisiae ( CAA44015 . 1 ) ; and Ustilago maydis ( XP_756473 ) . The accession numbers for the ACT1 and GPD1 genes are XP_566845 and XP_571627 , respectively . The sequence for C . neoformans var grubii ( CNAG_02001 . 1 [homologue in C . neoformans var . neoformans JEC21 , CNK03430] ) is from the Broad Institute database for C . neoformans var . grubii ( http://www . broad . mit . edu/annotation/genome/cryptococcus_neoformans/GeneIndex . html ) . | The ability of pathogens to regulate the export of proteins and other macromolecules is an important aspect of the infection process . The fungal pathogen Cryptococcus neoformans causes life-threatening infections in individuals with AIDS and delivers several virulence factors to the cell surface . These factors include polysaccharide material that forms a prominent capsule as well as the enzyme laccase that produces a protective layer of melanin in the cell wall . The cyclic adenosine 5′-monophosphate ( cAMP ) signaling pathway in C . neoformans plays a key role in sensing conditions such as nutrient availability to control expression of virulence factors , and defects in the pathway lead to attenuated or accentuated disease . Transcriptional profiling identified a regulatory link between the cAMP pathway and components of the machinery for transport to the cell surface . Studies with secretion inhibitors and with gene disruption mutants further supported connections between cAMP signaling , export functions , and the delivery of capsule and protein cargo outside the cell . These studies indicate that C . neoformans is a useful model for studying the regulation of secretion because of its particular dependence on this process for infection . In general , this work highlights the fact that components of the secretion machinery represent attractive targets for therapeutic measures to control fungal and other diseases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
]
| [
"yeast",
"and",
"fungi",
"infectious",
"diseases",
"genetics",
"and",
"genomics",
"microbiology"
]
| 2007 | Transcriptional Regulation by Protein Kinase A in Cryptococcus neoformans |
Visceral leishmaniasis is the most severe form of leishmaniasis . Worldwide , approximately 20% of zoonotic human visceral leishmaniasis is caused by Leishmania infantum , also known as Leishmania chagasi in Latin America . Current diagnostic methods are not accurate enough to identify Leishmania-infected animals and may compromise the effectiveness of disease control . Therefore , we aimed to produce and test two recombinant multiepitope proteins as a means to improve and increase accuracy in the diagnosis of canine visceral leishmaniasis ( CVL ) . Ten antigenic peptides were identified by CVL ELISA in previous work . In the current proposal , the coding sequences of these ten peptides were assembled into a synthetic gene . Furthermore , other twenty peptides were selected from work by our group where good B and T cell epitopes were mapped . The coding sequences of these peptides were also assembled into a synthetic gene . Both genes have been cloned and expressed in Escherichia coli , producing two multiepitope recombinant proteins , PQ10 and PQ20 . These antigens have been used in CVL ELISA and were able to identify asymptomatic dogs ( 80% ) more effectively than EIE-LVC kit , produced by Bio-Manguinhos ( 0% ) and DPP kit ( 10% ) . Moreover , our recombinant proteins presented an early detection ( before PCR ) of infected dogs , with positivities ranging from 23% to 65% , depending on the phase of infection in which sera were acquired . Our study shows that ELISA using the multiepitope proteins PQ10 and PQ20 has great potential in early CVL diagnosis . The use of these proteins in other methodologies , such as immunochromatographic tests , could be beneficial mainly for the detection of asymptomatic dogs .
Visceral leishmaniasis is caused by a protozoan parasite and affects approximately 500 , 000 million individuals annually worldwide [1] . Dogs are the main domestic reservoir of the causative agent of zoonotic visceral leishmaniasis , Leishmania infantum ( = L . chagasi ) [2] , [3] . Measures employed to control visceral leishmaniasis in Brazil since the 80's involve the elimination of infected dogs , among other actions [4] . Many serological tests have been described to detect canine anti-leishmanial antibodies [5]–[7] . However , current diagnostic tests lack sufficient sensitivity or specificity , including those recommended by the Brazilian Ministry of Health , Dual-Path Platform ( DPP; Bio-Manguinhos/Fiocruz , Rio de Janeiro , Brazil ) and Canine Leishmaniasis ELISA kit ( EIE-LVC kit; Bio-Manguinhos/Fiocruz , Rio de Janeiro , Brazil ) [8] , [9] . Thus , canine visceral leishmaniasis ( CVL ) represents a serious public health issue given the frequency of asymptomatic infections ( up to 85% ) [10] , and given that both asymptomatic and symptomatic dogs are equally infectious to the vectors [11] , [12] . Therefore , the development of accurate diagnostic methods for canine infection , mainly for asymptomatic animals , is essential for visceral leishmaniasis surveillance programs , in addition to understanding immunological responses in resistant or susceptible animals . Recently , an increasing number of Leishmania antigens have been evaluated in serodiagnosis [8] , [13]–[16] . High values of sensitivity and specificity are very important to these antigens . However , if the objective is a screening test , high sensitivity is desirable . But if a confirmatory test is being developed , high specificity becomes more important in this case . In an attempt to achieve high sensitivities and specificities in tests , an alternative approach is the use of multiepitope proteins , which have been demonstrated to be a valuable tool in CVL diagnosis . Soto et al . [17] evaluated a chimeric protein for the diagnosis of L . infantum-infected ( n = 59 ) and uninfected dogs ( n = 15 ) and showed 79% of sensitivity and 96% of specificity . Boarino et al . [18] tested another chimeric antigen in 232 animals in which leishmanial infection had been detected by parasitological examination ( n = 19 ) or RIFI ( n = 213 ) . They showed that this chimeric antigen had 96% of sensitivity and 99% of specificity in this diagnosis . Despite these promising results , these antigens were tested only in the Old World . Therefore , tests with other multiepitope proteins are still needed , not only in other geographical areas , but also in dogs with different clinical status , in comparison to molecular assays and in follow up of canine infection . Early serological detection of CVL is highly desirable , mainly because in urban areas there is a high prevalence of infected dogs ( as determined by PCR ) that are not detected by conventional serological methods [19] . Therefore , we aimed to search for new antigens that could be used to detect CVL in an early phase . In this study , we constructed two multiepitope proteins using epitopes previously identified [20] . The protein named PQ20 , constituted by twenty peptides , presented approximately 95% of B cell epitopes when mapped with BCPreds and ABCPred programs . The other protein , named PQ10 , was constituted by ten antigenic peptides that showed good results in CVL ELISA , with accuracy up to 0 . 94 . Some of them gave positive reactions in up to 95% of asymptomatic dog sera . When these peptides were mixed into a single solution , good results were also obtained [8] . These two multiepitope proteins were tested in asymptomatic dog detection and also in early detection ( before PCR positive results ) of CVL . In this work , the main contribution of multiepitope antigens would be the usefulness in canine management , due to their ability to detected infected animals in a serological test , with similar sensitivitiy to PCR .
Serological tests were performed in 3 steps: 1 ) First , we used a sera panel ( n = 52 ) in ELISA with the multiepitope proteins , tested in parallel with DPP and EIE-LVC kit; 2 ) To validate ELISA with the multiepitope proteins , we used a multicentric panel with 131 serum samples ( negative in indirect immunofluorescence reaction - RIFI and EIE-LVC kit ) and 231 serum samples that were positive in RIFI and EIE-LVC kit; 3 ) A panel of 42 serum samples was tested in ELISA with the multiepitope proteins; these sera were acquired from dogs which were followed for 18 months and were initially negative in PCR and EIE-LVC kit . In the first step , the uninfected dogs ( n = 9 ) were negative based on parasitological and serological tests ( RIFI and EIE-LVC kit ) , while infected animals ( n = 43 ) were certified by parasitological tests conducted on bone marrow cells examined by optical microscopy , and also by RIFI and EIE-LVC kit . Within samples of infected dogs , n = 33 were clinically undefined while n = 10 were asymptomatic according to typical CVL signs , stated elsewhere [21] . In the second step , ELISA validation , we used a multicentric panel with 362 samples , of which n = 131 negative in RIFI and EIE-LVC kit . Within this negative group , n = 40 serum samples were from animals born in kennels with wire mesh in a non-endemic area ( Ouro Preto , Minas Gerais , Brazil; Southeast ) , being the N1 subgroup . Then , n = 91 serum samples were from endemic area animals ( Belo Horizonte , Minas Gerais , Brazil; Southeast ) . Within these samples , n = 22 had negative results by PCR , being the N2 subgroup; the others ( n = 69 ) , which were not tested by PCR , were classified as N3 subgroup . Conversely , n = 231 serum samples had positive results in RIFI and EIE-LVC kit , and were clinically undefined . Within this group , n = 87 samples were from endemic area animals ( Teresina , Piauí , Brazil; Northeast ) and were also positive in the DPP test , being the P1 subgroup . The other samples ( n = 144 ) were from dogs living in another endemic area ( Belo Horizonte , Minas Gerais , Brazil ) . Among them , n = 22 samples were also positive by PCR , being the P2 subgroup; the other n = 122 samples were not tested by PCR , being the P3 subgroup ( Table 1 ) . Finally ( third step ) , samples from a greater cohort study [22] were included in this work . Canine blood was collected in three different periods: in the beginning of the cohort ( T0 ) , 12 months later , and 18 months later . During all this time , they were maintained in their households , exposed to all the risk factors . Sera from these dogs ( n = 42 ) were tested by PCR-RFLP , EIE-LVC kit , as previously described [22] , and ELISA with multiepitope proteins ( PQ10 and PQ20 ) . All animals were from Belo Horizonte , Minas Gerais state , Brazil , and were followed to verify seroconversion in this endemic area . All serum samples came from already existing collections , from Laboratório de Leishmanioses of Universidade Federal de Minas Gerais , Minas Gerais state , Brazil and from Laboratório de Imunoparasitologia of Universidade Federal de Ouro Preto , Minas Gerais state , Brazil . Two different multiepitope synthetic genes were designed . First , we joined 10 coding sequences from antigenic peptides [8] , resulting in PQ10 . Second , we joined 20 coding sequences from T and B cell epitopes [20] , resulting in PQ20 . A flexible linker ( Gly-Ser-Gly-Ser-Gly ) coding sequence was used as a spacer between epitope sequences [23] . NdeI and NotI restriction sites were added to the 5′ and 3′ ends , respectively , of both synthetic genes to aid in cloning . A 6xHIS tag coding sequence was added upstream of the stop codon of each synthetic gene for affinity purification of recombinant proteins . Both sequences were codon-optimized for Escherichia coli expression . PQ10 and PQ20 genes were commercially synthesized by Genscript , USA . Synthetic genes were cloned into the NdeI and NotI restriction sites of a pET9a24a expression vector , resulting in pET-PQ10 and pET-PQ20 . Sequence analysis of the cloned fragments confirmed the correct fusion and orientation of the insert . Recombinant plasmids were used to transform E . coli C41 strain and protein expression was carried out by inoculating 500 mL of Luria Bertani medium containing 0 . 05 mg/mL kanamycin with an overnight bacterial culture . This suspension was incubated on a rotary shaker ( 200 rpm ) at 37°C until an optical density of 0 . 6 at 600 nm . Protein expression was induced with 0 . 4 mM IPTG ( isopropyl-β-D-thiogalactopyranoside ) for 4 h on a rotary shaker ( 200 rpm ) at 37°C . Cells were lysed using a microfluidizer ( EmulsiFlex C3 , Avestin ) and soluble and insoluble protein fractions were analyzed by SDS-PAGE [24] . Next , insoluble fractions of recombinant proteins were affinity purified using an ÄKTA Prime chromatography system ( GE Healthcare Life Science ) with a 5 mL HIS-Trap FF column ( GE Healthcare Life Science ) , in the presence of 8 M urea , according to manufacturer's instructions . To evaluate the antigenicity of multiepitope proteins , ELISA was conducted with both PQ10 and PQ20 . All ELISA procedures were optimized in terms of antigen concentrations , dilutions of serum and conjugated immunoglobulins , and the microplates that would be employed . Falcon flexible microplates ( Becton Dickinson® , France ) for PQ10 and Eppendorf microplates ( Hamburg , Germany ) for PQ20 were coated for 16 h approximately with 0 . 5 µg/mL recombinant proteins diluted in 0 . 05 M carbonate buffer ( pH 9 . 6 ) at 4°C . After three washes with PBS/T ( PBS: 10 . 14 mM Na2HPO4; 1 . 37 mM KH2PO4; 146 mM NaCl; 2 . 64 mM KCl , pH 7 . 4 , containing 0 . 05% Tween20 ) , wells were blocked with 5% powdered skimmed milk in PBS/T at 37°C for 1 h . Serum samples , diluted 1∶200 in PBS/T containing 0 . 5% powdered skimmed milk , were added and incubated at 37°C for 1 h . After three washes , plates were incubated with peroxidase-conjugated anti-dog immunoglobulin G ( Sigma , reference A9042 ) , diluted 1∶2500 in PBS/T containing 0 . 5% powdered skimmed milk at 37°C for 1 h . After washing three times , reactions were developed with Fast-OPD ( Sigma ) and plates were incubated for 30 min in the dark , following manufacturer's instructions . Reactions were stopped with 2 M H2SO4 , and plates were read at 492 nm in a MultiskanGo ( ThermoScientific ) plate reader . Samples that provided discordant results ( positive in RIFI and EIE-LVC kit but negative in ELISA with PQ10 and PQ20 ) were tested by another methodology . The immunochromatographic test Kalazar Detect ( InBios ) was used according to manufacturer's instructions . Additionally , anti-CVL antibody concentration in serum samples was titrated by RIFI [25] . Recombinant proteins sensitivity and specificity were calculated using parasitological results as a gold standard . A cut off point for optimal sensitivity and specificity was determined using ROC analysis [26] , and the area under the curve ( AUC ) was calculated to assess performance of the tests . Once the intended use of multiepitope proteins is in both screening and confirmatory tests , the highest values of sensitivity and specificity were desirable . All of the statistical analyses were performed using GraphPad Prism ( version 5 . 0 ) .
The two multiepitope proteins , PQ10 and PQ20 , were successfully produced by E . coli C41 strain under the conditions described in Methods . Amino acid sequences ( with flexible linkers ) are shown in Fig . 1 . Fig . 2 shows SDS-PAGE analysis of highly expressed recombinant proteins at the expected sizes . A band of approximately 21 kDa , referring to PQ10 , and another of approximately 33 kDa , referring to PQ20 , were observed . Both recombinant proteins were found mainly in the insoluble fraction of the cell lysate . Fig . 2 also shows fractions of the purified multiepitope proteins . When comparing the four different CVL diagnostic tests: 1 ) DPP; 2 ) EIE-LVC kit; 3 ) ELISA with PQ10; and 4 ) ELISA with PQ20 , large differences were found . These tests were performed with sera from uninfected ( n = 9 ) and infected dogs ( n = 43 ) , as described in Methods . While sensitivities of EIE-LVC kit and DPP were 64 . 5% and 72 . 9% respectively , ELISA-PQ10 and ELISA-PQ20 showed sensitivities of 88 . 8% and 84 . 9% , respectively ( Table 2 ) . However , ELISA-PQ10 and ELISA-PQ20 ( ELISA-PQs ) specificities were lower ( PQ-10: 80% and PQ-20: 65% ) when compared to other test specificities ( DPP: 90% and EIE-LVC kit: 100% ) . Additionally , multiepitope protein PQ10 was able to detect 80% of asymptomatic infected dogs , which was not observed with EIE-LVC kit . This test showed no asymptomatic dog detection while DPP detected only 10% of them . Validation was performed with 231 serum samples from L . infantum-infected dogs and 131 serum samples from uninfected dogs as the negative controls . Promising results were achieved with ELISA-PQs , which provided global sensitivities of 80 . 2% ( PQ10 ) and 84 . 9% ( PQ20 ) . The specificity was the same for both , 65 . 6% . According to Swets [27] , these antigens provided tests with moderate accuracy ( AUC = 0 . 82 for PQ10 and AUC = 0 . 84 for PQ20 ) . In Fig . 3 and 4 , the absorbance distribution of ELISA-PQ10 and ELISA-PQ20 , respectively , is shown , as well as the cut off point for optimal sensitivity and specificity . Analyzing the different subgroups , we observed that between those with only two serological concordant tests ( RIFI and EIE-LVC kit ) , N3 and P3 , specificities are lower when compared to those groups , N2 and P2 , with an additionally parasitological concordant test , as PCR ( Table 3 ) . Specificities ranged from 65 . 5% to 71 . 0% when comparing subgroups N3 and P3 and from 90 . 9 to 72 . 7 in the other comparison ( N2 and P2 ) . We observed that a third methodology to guarantee CVL diagnosis ( born in kennels with wire mesh or an additional concordant test such as DPP or PCR ) decreases the number of false diagnosis by PQs . Furthermore , it was possible to notice that ELISA-PQs' performance were similar when testing samples from different places . Samples from groups P1 ( from Teresina , Piauí state , Brazil; Northeast ) and P3 ( from Belo Horizonte , Minas Gerais state , Brazil; Southeast ) for example , showed similar results , indicating that the antigens present satisfactory performance in different CVL endemic areas . Moreover , besides the fact that ELISA-PQs presented greater sensitivities and specificities in groups with concordant parasitological and serological results , these results also indicate that these novel antigens could lead to improvements of diagnostic tests . To investigate the occurrence of false-negative results in ELISA-PQs , samples that showed discordant results were tested by Kalazar Detect . This method provided only 56% of positivity when testing samples that were negative in ELISA-PQs and positive in EIE-LVC and RIFI . Thus , to test if these samples had low titers of anti-Leishmania antibodies , a quantitative RIFI was performed , until a maximum titer of 1∶2560 . Indeed , the majority of samples had low titers of antibodies , with only 9 . 5% of them ( from n = 63 ) possessing titers above 1∶320 . It showed that ELISA-PQs cannot easily react with serum samples that have low antibody titers . Samples from n = 42 CVL endemic area dogs ( negative in PCR and EIE-LVC kit ) were analyzed by ELISA-PQs and were followed for 18 months . In the first analysis ( T0 ) , ELISA-PQ10 already identified 51 . 1% of the samples as positive and ELISA-PQ20 , 65 . 1% ( Fig . 5 ) . Afterwards , PCR provided positive results , reaching a positivity of 86 . 0% after 18 months . These positivities were not accompanied by EIE-LVC kit , which detected only 4 . 6% of the cases starting at the second analysis ( after 12 months ) . Recombinant proteins were able to detect positive dogs starting at the first analysis , detecting a similar number of cases after 12 months , with a decrease after 18 months . This suggests that the antibodies detected by ELISA-PQs reach maximum levels just after infection , decreasing thereafter . Besides , all animals which provided ELISA PQs' positive results in T0 also provided PCR positive results after 12 months , suggesting early detection by multiepitope proteins .
Serological methods are powerful tools in CVL diagnosis , being frequently used for canine mass screening . Several Leishmania antigens have been characterized , and recombinant technology has been used for the development of new enzymatic immunoassays [16] , [28] , [29] . Given the lack of sensitivity of Brazilian recommended assays to CVL diagnosis , mainly in asymptomatic dogs [8] , [9] , the search for new antigens is still needed . In this work , we have produced and validated two new recombinant antigens for CVL diagnosis . They are multiepitope proteins , resulting from synthetic gene design , which has been referred elsewhere as a useful methodology [30] , [31] . The validation of immunossorbent assays with our recombinant proteins was a very important contribution , because many antigens have not been tested in multicentric studies with a large panel of samples , mainly from uninfected animals , as we performed . Other multiepitope proteins have already been used for the diagnosis of Leishmania-infected dogs , showing promising results [17] , [18] . However , this is the first time that serological results are similar to those of parasitology , providing an early serological Leishmania diagnosis . Our results indicate that the main usefulness of ELISA-PQs is in serological screening tests , once sensitivities exceeded 80% . However , specificities were about 66% , indicating a need for other confirmatory tests with better specificity . As stated elsewhere [32] , the use of only one antigen in screening and confirmation tests in CVL diagnosis increases the number of false-positives . Furthermore , our multiepitope proteins could be useful in canine management due to their ability to detect infected animals with similar sensitivity to PCR , in a serological assay . Main advantages of using these multiepitope proteins are low cost and easy automation in ELISA tests . ELISA-PQs showed moderate specificities with samples that had two concordant serological tests . This criterion of sample election is adopted by many authors [33]–[35] . However , it could have included infected dogs ( before seroconversion ) in the group of uninfected animals that were not detected by EIE-LVC kit and RIFI . On the other hand , our antigens are able to detect CVL infection before serological tests , as well as a molecular test ( PCR ) ; hence , the false-positive results could be correct ( not false ) . Another explanation regarding low specificity could be the occurrence of cross reactions , leading to positive results in group N1 ( dogs born in kennels with wire mesh in a non-endemic area dogs ) . Besides , when analyzing the false-negative results from ELISA-PQs , we observed that most of these samples ( 90 . 5% ) presented titers lower than 1∶320 , suggesting that our multiepitope proteins have difficulties in reacting with low titer samples . Similar to our results , Reis et al . [34] showed 12 . 5% of false-negative results when testing ELISA-k39 , due to the presence of dogs with antibody titers lower than 1∶320 . Other authors [14] have showed 15% of false-negative results also due to low antibody titers , in a latex agglutination test with the A2 antigen . Detection of asymptomatic dogs in ELISA-PQs ( 80% of tested samples ) was substantially superior compared to the detection by DPP or EIE-LVC kit . This is a relevant issue regarding visceral leishmaniasis control , as asymptomatic dogs play an important role in the epidemiological cycle of the disease [11] . Similar to our data , Martins et al . [16] also showed promising results with the recombinant protein LiHyp1 , which detected 94% of asymptomatic dogs . In high transmission areas of visceral leishmaniasis , with prevalence above 3% , asymptomatic dogs must be included in control measures to guarantee program success [36] . Prevalence of CVL is very changeable in different areas in the world and also in different regions of the same country , as observed in Brazil . So , the use of a sorological test should consider this prevalence once it influences in negative and positive predictive values of a test . Although in low prevalence our proteins could not be the most useful antigens , they are still helpful tools to detect early infection . An important point relative to the detection of asymptomatic dogs is the ability to follow up on resistant animals . Thus , the described proteins could be used in the future to detect and follow up canine infection . Our multiepitope proteins were able to detect circulating antibodies in the early phase of infection . All serum samples which were positive in ELISA-PQs were also positive in PCR thereafter . Positivities detected by ELISA-PQs , as well as by PCR , were not accompanied by EIE-LVC kit . As already stated , PCR provides positive results before serological tests [21] , [37] . However , immunoassays with PQs apparently detect other types of immunoglobulins . One such case could be detection of IgMs , which are rapidly and transiently produced in the initial stage of CVL infection . This immunoglobulin could have been detected by the enzyme-conjugated antibody used in our ELISA-PQ tests ( anti whole molecule ) . Some authors [38] have shown that canine IgG levels increase from the second month after experimental infection . Theoretically , from this point on , it is possible to serologically detect the infection , which occurred with ELISA-PQs . However , few studies are found in the literature about serological early detection of L . infantum infection . Quinnell et al . [21] showed that serology failed to detect 43 of 343 ( 12 . 5% ) confirmed post-infection samples; 36 of these were from dogs early in infection ( prior to seroconversion ) . In an experimental work [39] , Leishmania infection was detected by ELISA with crude antigen in 37% of the dogs after 90 days . Similarly , Rodríguez-Cortés et al . [40] have detected 33 . 3% of dogs the same time post-infection . Faria et al . [8] have described the peptides that compose PQ10 structure . When these synthetic peptides were tested together in ELISA , mixed in a single solution , the sensitivity was 75% and specificity , 95% in canine diagnosis . In comparison to multiepitope protein PQ10 , an increase in sensitivity ( 88 . 8% ) and also a decrease in specificity ( 80% ) were observed . Thus , depending upon the use of the antigen , multiepitope protein or mixed peptides could be chosen , in order to have higher sensitivities or specificities . Recent studies have evaluated the development of immunochromatographic tests with recombinant antigens to detect different pathologies [41]–[43] , representing a possible approach for our proteins . These recombinant multiepitope antigens could be combined in an attempt to improve accuracy . Ideally , it would be interesting to use an antigen with good sensitivity , as observed with our multiepitope proteins , and another antigen with good specificity in the same strip test . In conclusion , we have designed two new multiepitope recombinant proteins to improve CVL serodiagnosis . Our findings indicate that PQ10 and PQ20 could be useful for serodiagnosis and allow for the detection of asymptomatic dogs , as well as dogs in early phase of infection . The development of an immunochromatographic test using these proteins would be a valuable tool for CVL diagnosis . | Visceral leishmaniasis is the most severe form among leishmaniasis , being a neglected disease caused by a protozoan parasite . Its transmission through phlebotominae bites , between dogs and humans , classifies it as a zoonotic disease . It is caused by the specie Leishmania infantum ( = L . chagasi ) and represents 20% of the world's human visceral leishmaniasis . Visceral leishmaniasis is a serious public health issue , fatal if untreated , and its incidence is increasing in urban areas of the tropics . In Brazil , one of the control measures is the identification and elimination of infected dogs , which act as reservoirs for Leishmania parasites . Diagnostic methods used to identify infection in these animals are still not accurate enough , which may compromise the effectiveness of this control measure . Thus , to contribute to the diagnosis of canine visceral leishmaniasis , we aimed to develop and test two new antigens that could be applied in early detection of infected dogs . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
]
| [
"gene",
"expression",
"and",
"vector",
"techniques",
"biochemistry",
"veterinary",
"parasitology",
"proteins",
"biology",
"and",
"life",
"sciences",
"molecular",
"biology",
"techniques",
"recombinant",
"proteins",
"molecular",
"biology",
"parasitology",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques"
]
| 2015 | Novel Recombinant Multiepitope Proteins for the Diagnosis of Asymptomatic Leishmania infantum-Infected Dogs |
Knowledge of the molecular and genetic mechanisms underlying the separation of dendritic and axonal compartments is not only crucial for understanding the assembly of neural circuits , but also for developing strategies to correct defective dendrites or axons in diseases with subcellular precision . Previous studies have uncovered regulators dedicated to either dendritic or axonal growth . Here we investigate a novel regulatory mechanism that differentially directs dendritic and axonal growth within the same neuron in vivo . We find that the dual leucine zipper kinase ( DLK ) signaling pathway in Drosophila , which consists of Highwire and Wallenda and controls axonal growth , regeneration , and degeneration , is also involved in dendritic growth in vivo . Highwire , an evolutionarily conserved E3 ubiquitin ligase , restrains axonal growth but acts as a positive regulator for dendritic growth in class IV dendritic arborization neurons in the larva . While both the axonal and dendritic functions of highwire require the DLK kinase Wallenda , these two functions diverge through two downstream transcription factors , Fos and Knot , which mediate the axonal and dendritic regulation , respectively . This study not only reveals a previously unknown function of the conserved DLK pathway in controlling dendrite development , but also provides a novel paradigm for understanding how neuronal compartmentalization and the diversity of neuronal morphology are achieved .
The separation of the dendritic and axonal compartments in neurons is prerequisite to the function of neural circuits . Although the difference between dendrites and axons is a cornerstone of modern neuroscience , as theorized in the “neuron doctrine” by Ramon y Cajal [1] , our molecular understanding of how neuronal compartmentalization is achieved remains limited . This knowledge , however , is crucial for understanding the assembly of neural circuits . Moreover , it is needed to develop strategies that will correct defective dendrites or axons with subcellular precision , and to alter the wiring of neural circuits in animal models in order to interrogate the functions of the nervous system . Previous studies have demonstrated the existence of regulators dedicated to dendrite or axon growth in the same neuron , referred to as “dedicated mechanisms” herein . For instance , the transcription complex , p300–SnoN , specifically promotes axon growth in the cerebellar granule neurons [2] . In contrast , transcription factor NeuroD is dedicated to dendritic growth in mammalian cerebellar granule neurons [3] . Likewise , bone morphogenetic protein 7 ( BMP7 ) specifically promotes dendritic growth in several types of neurons in culture [4] , [5] . In Drosophila , the transcription factor Dendritic arbor reduction 1 ( Dar1 ) promotes dendritic , but not axonal , growth [6] . In addition , dendritic and axonal growth exhibit differences in their dependence on the secretory pathway [7] . Besides the dedicated mechanisms , another way to differentiate dendritic and axonal growth is through bimodal regulators that control dendritic and axonal growth in opposite directions [8]–[11] . Different from the dedicated mechanisms , the bimodal mechanisms may coordinate growth of the two neuronal compartments . However , how the function of a molecule or signaling pathway diverges into distinct dendritic and axonal regulations is poorly known . In this study we report that the dual leucine zipper kinase ( DLK ) signaling pathway is a novel bimodal regulator for dendritic and axonal growth in vivo . The core players in the DLK signaling pathway are the DLK and the Pam/Highwire/RPM-1 ( PHR ) family of E3 ubiquitin ligases that suppress DLK expression . The PHR-DLK signaling module plays an important role in axon development , as demonstrated by studies in C . elegans [12]–[14] , Drosophila [15]–[17] , zebrafish [18] , [19] , and mammals [20]–[22] . Loss of the Drosophila homologue of DLK-1 , Wallenda ( Wnd ) , suppresses the axonal overgrowth caused by loss of the PHR protein Highwire ( Hiw ) [15] , [17] . Consistently , overexpression of Wnd promotes axonal growth of motoneurons in Drosophila larvae [17] . In Drosophila adult mushroom body neurons , Hiw-Wnd pathway is required for the segregation of axon branches in response to guidance cues [23] . In addition to the roles in axon development , recent studies have discovered a conserved function of the DLK pathway in axon regeneration [24]–[28] and degeneration in several species [29]–[32] . Although these exciting findings have established critical roles for the DLK pathway in axon development , regeneration , and degeneration , whether the DLK pathway regulates dendrites remains unknown . Here we show that the DLK pathway directs the growth of axons and dendrites in opposite directions in the class IV dendritic arborization ( C4da ) neurons in Drosophila . By inhibiting Wnd functions , Hiw restricts axonal growth but promotes dendritic growth . The opposite effects of the Hiw-Wnd pathway on axons and dendrites are achieved through two distinct transcription factors: Fos , which mediates the regulation of axonal growth , and Knot ( Kn ) , which mediates the regulation of dendritic growth . Collectively , these results demonstrate that a single signaling pathway can differentiate dendritic and axonal growth through two independent transcriptional programs .
All functional studies of the PHR-DLK pathway in neurons have so far focused on axons . We first set out to determine whether the PHR gene hiw is involved in dendrite development using Drosophila as a model system . The C4da neurons in Drosophila larva are a well-established in vivo system for studying the molecular mechanisms of dendrite and axon development . The dendrites and axons of these neurons are distinguishable from each other at both molecular and organelle levels in a way that resembles mammalian neurons [33] . Moreover , these neurons are amenable to single-cell genetic manipulations [33] , [34] , which is important for comparing dendritic and axonal development in vivo . In each hemi-segment of a larva , there are three C4da neurons ( ddaC , v'ada , and vdaB ) , whose cell bodies are located respectively in the dorsal , lateral , and ventral parts of the body wall . The axons of the three C4da neurons extend to the ventral nerve cord ( VNC ) where the terminals form a ladder structure ( Figure S1A ) . At single-cell resolution , the axon terminal of each C4da neuron consists of an anterior projection that extends within one segment length . ddaC and vdaB neurons also extend a contra-lateral branch and sometimes a posterior branch ( Figure S1A′ ) [35] . Collectively , the axon terminals of the three C4da neurons form a fascicle that connects two adjacent neuropils ( Figure S1A′ ) . To examine the role of hiw in dendritic development , we labeled the C4da neurons in hiw mutant larvae using a C4da-specific marker , ppk-CD4::tdTomato [34] , [36] . We found that dendritic growth was dramatically reduced in the null allele hiwΔN and , to a lesser extent , in the hypomorphic hiwND8 mutants ( Figure 1A and B ) . Both total length and number of termini of dendrites were significantly reduced in hiwΔN and hiwND8 mutants ( Figure 1B ) . Consistent with the known function of hiw in suppressing axonal growth [15] , [16] , hiw mutations led to exuberant growth of axon terminals in C4da neurons . In hiw mutant larvae , thickened connective fascicles were observed in the C4da neuropil ladder ( Figure S1B ) . In wild-type larvae , there was no hemi-segment that contained more than three longitudinal connectives between the axon entry point of abdominal segment 5 ( A5 ) and that of A6 ( Figure S1A′ , B , and D ) . In contrast , 100% of hiw mutant C4da neuropils exhibited more than three connectives ( Figure S1B and D ) , which could either arise from an increased number of axon branches from neurons in the same segment or from overextended axons that normally remain in other segments . Our further analysis showed that the effects of hiw mutations on dendritic and axonal growth are not a result of defective dendrite and axon identities . The axon-specific marker , Kinesin-β-galactosidase [37] , [38] , remained exclusively localized to the axons of C4da neurons that were mutant for hiw ( Figure S2A ) . Furthermore , the initial growth and pathfinding of axons to the VNC or the extension of minor dendritic processes remained unaltered in embryos devoid of both maternal and zygotic hiw ( Figure S2B ) . Thus , hiw appears to be dispensable for early development , including the initial specification of axon and dendrite . Taken together , these results suggest that hiw plays a dichotomous role in differentiating dendrite and axon growth after their identities have been specified . Previous studies of axon development have discovered both cell-autonomous [16] , [17] and non-cell-autonomous roles of hiw [23] . To determine whether hiw functions cell-autonomously in C4da neurons and to examine the axon and dendrite defects at single-neuron resolution , we generated hiw mutant neurons with the Mosaic Analysis with a Repressible Cell Marker ( MARCM ) technique [39] . Consistent with the reduced dendritic growth in hiw mutant larvae , we observed a reduction of high-order dendritic branches in hiw loss-of-function mutant neurons ( Figure 1C ) . Moreover , fewer dendritic branches arrived at the segment border as compared to wild-type . hiw mutations caused a 43% reduction in total dendrite length and 40% reduction of the number of dendrite termini ( Figure 1E ) . In contrast to their dendritic defects , hiw mutations resulted in a 2 . 4-fold increase of axon terminal length ( Figure 1D and F ) as compared to wild-type . The axon terminals of hiw mutant neurons typically spanned multiple segments , whereas the vast majority of wild-type C4da neurons extended axonal branches between their own segments and the anterior neighboring segments ( Figure 1D ) . Noticeably , although the axon terminals of hiw mutant neurons grew exuberantly , they preserved normal guidance within the C4da neuropil tracts . In agreement with the MARCM results , overexpressing Hiw in C4da neurons rescued both dendritic and axonal defects in hiw mutant larvae to a level comparable to wild-type ( Figure S3 ) , further confirming that the loss of hiw in C4da neurons is responsible for the dendritic and axonal defects . Overexpression of Hiw alone did not significantly alter axonal or dendritic growth ( Figure S3 ) , suggesting that hiw is necessary but insufficient to instruct dendritic growth and restrict axon growth . Taken together , these results demonstrate that Hiw functions as a cell-intrinsic bimodal regulator of dendritic and axonal growth in C4da neurons . Two parallel downstream pathways are known to mediate axon overgrowth induced by loss of PHR proteins . First , the PHR orthologs in C . elegans ( rpm-1 ) and Drosophila ( hiw ) suppresses the worm dlk-1 and the fly DLK wallenda ( wnd ) , respectively , to restrain axonal growth in motoneurons [14] , [17] . Second , the worm rpm-1 regulates a trafficking pathway that consists of the Rab guanine nucleotide exchange factor ( GEF ) GLO-4 and the Rab GTPase GLO-1 , which restrict axon extension in mechanosensory neurons and synaptogenesis in motoneurons [40] . In order to delineate the mechanism underlying the bimodal control of dendritic and axonal growth by hiw , we tested the involvement of these two pathways in axon and dendrite growth in C4da neurons . While wnd loss-of-function mutations on their own did not alter axonal ( Figure S1C–D ) or dendritic morphology ( Figure 2 ) , they completely suppressed both axonal and dendritic defects caused by hiw mutations ( Figure S1C–D and Figure 2 ) . These observations suggest that wnd acts downstream of hiw to promote axonal growth and inhibit dendritic growth . Consistent with this model , overexpression of Wnd in C4da neurons induced extensive axon terminal overgrowth and profoundly reduced dendritic branching in C4da neurons ( Figure S1C–D and Figure 2 ) . In contrast , overexpression of a kinase-dead ( KD ) form of Wnd resulted in morphologically normal C4da neurons ( Figure S1C–D and Figure 2 ) . Hence , increased expression of the Wnd kinase is sufficient to inhibit dendritic growth and promote axonal growth . We also examined the potential involvement of the Rab trafficking pathway by testing Drosophila homologs of glo-4 and glo-1 in axon and dendrite development in C4da neurons . In C . elegans , glo-4 mutants exhibited axon overextension similar to that in rpm-1 mutants [40] . Overexpressing the Rab GTPase Glo-1 , which is activated by Glo-4 , partially rescued axon termination defects in rpm-1 mutants [40] . The Drosophila homologs of glo-4 and glo-1 are claret ( ca ) and lightoid ( ltd ) , respectively [41] . The ca mutant MARCM clones devoid of maternal contribution exhibited axons and dendrites that were indistinguishable from wild-type clones ( Figure S4A–D ) . In addition , overexpressing Ltd failed to rescue either axon or dendrite defects in hiw mutants ( Figure S4E–H ) . These results suggest that Drosophila C4da neurons use the DLK ( Wnd ) pathway , rather than the Ca-Ltd vesicle trafficking pathway , to mediate hiw function in axonal and dendritic growth . How might the Hiw-Wnd pathway control axonal and dendritic growth differently in the same neurons ? In Drosophila motoneurons , the Hiw-Wnd pathway requires the transcription factor Fos [17] . Fos is phosphorylated by Bsk ( JNK ) [42] , which positions it as the downstream kinase of the Wnd-Hep7-JNK kinase cascade [17] . Overexpressing a dominant negative form of Fos partially suppresses axonal overgrowth at the NMJ of hiw mutants [17] . Because of this , we decided to examine whether Fos is required by Wnd to promote axonal growth in C4da neurons . To test the role of Fos with loss-of-function mutants , and to bypass lethality caused by fos null mutations kay1 [43] , [44] , we generated kay1 MARCM clones in the presence or absence of a UAS-Wnd transgene that overexpresses Wnd ( OE Wnd ) . kay1 alone did not alter axonal growth ( Figure 3A ) , but completely suppressed the axon overextension caused by Wnd overexpression ( Figure 3A and C ) , which suggests that fos is required for Wnd-induced axonal overgrowth . In contrast to the axonal role of Fos , kay1 did not block the dendritic reduction caused by Wnd overexpression . The total dendritic length of MARCM clones that overexpressed Wnd in the kay1 background ( OE Wnd+kay1 ) was indistinguishable from that of Wnd-overexpressing clones ( Figure 3B , B′ , and C ) , and the number of dendrite termini was further reduced from that of Wnd-overexpressing clones . Interestingly , the kay1 mutation alone caused a mild reduction in dendritic length and branch number ( Figure 3B , B′ , and C ) . This result suggests that , although Fos does not mediate the dendritic functions of the DLK pathway , it plays a minor role in supporting dendritic growth . Taken together , these results suggest that Wnd acts through Fos to specifically promote axonal growth . In order to understand how the function of DLK pathway diverges into dendritic and axonal regulations , we hypothesized that the divergence occurred at the transcriptional level , and therefore tested the transcription factors that are known to regulate dendritic growth in da neurons . Among them , the Krüppel-like factor Dar1 , the homeodomain transcription factor Cut ( Ct ) , and zinc-finger transcription factor Knot ( Kn , as known as Collier ) have been shown to be essential for dendritic growth in C4da neurons . Loss-of-function mutations in each of these transcription factors severely reduce dendritic growth in C4da neurons [6] , [45]–[48] . We first tested whether expression levels of these transcription factors in C4da neuron nucleus were altered in hiw loss-of-function mutants . No significant difference in the levels of Dar1 [6] or Cut [45] was observed between wild-type and hiw mutant C4da neurons ( Figure S5A–C ) . In contrast , the nuclear levels of Kn , which belongs to the evolutionarily conserved Collier/Olf1/EBF ( COE ) family , were significantly reduced in both hiw mutant neurons and Wnd-overexpressing neurons ( Figure 4A and B ) Kn is required for the expression of the ENaC ion channel Pickpocket ( Ppk ) in C4da neurons [46]–[48] . Kn loss-of-function mutations reduce ppk transcription [46] and suppress ppk promoter activity as assayed with a ppk-eGFP transgene ( Figure 4D ) [47] , [48] . Furthermore , misexpression of Kn induces ectopic ppk-eGFP expression in neuron types that do not normally express ppk-eGFP [46]–[48] . Therefore , the ppk-eGFP transgenes may be used as readout for Kn transcriptional activity . Consistent with the reduced Kn expression by hiw mutations or Wnd overexpression , we found a 37% reduction in ppk-eGFP fluorescence intensity in the soma of hiw mutant C4da neurons and a 68% reduction in those of Wnd-overexpressing neurons ( Figure 4C and D ) . Furthermore , overexpressing Kn rescued the reduced expression of ppk-eGFP in hiw mutant or Wnd-overexpressing neurons ( Figure 4C and D ) . The correlation between ppk-eGFP fluorescence intensity and Kn levels suggests that the Hiw-Wnd pathway controls Kn transcriptional activity by regulating its protein levels . Nevertheless , it does not rule out the possibility of posttranslational regulation of Kn activity by Hiw-Wnd . Taken together , Hiw suppresses Wnd function , thus maintaining high levels of Kn protein in C4da neurons , which is required for dendritic growth . It has been demonstrated that loss-of-function mutations of kn cause reduction in dendritic length and branch numbers [46]–[48] . We tested potential genetic interactions between hiw and kn in controlling dendritic growth . C4da dendrites developed normally in both hiwΔN/+ heterozygous and knKN4/+ heterozygous larvae ( Figure 5A and B ) , in which Kn expression and ppk-eGFP levels remained comparable to wild-type ( Figure S5D–F ) . In contrast , the hiwΔN/+; knKN4/+ transheterozygous larvae exhibited dramatically reduced dendritic growth ( Figure 5A–B ) , revealing a strong genetic interaction between hiw and kn . We investigated the nature of the genetic interaction by epistasis analysis . Kn overexpression resulted in a mild 16% reduction of C4da dendritic length ( Figure 5C–D ) , possibly due to destabilized microtubules as a result of increased expression of the microtubule severing protein Spastin [6] , [48] . Nevertheless , overexpressing Kn in hiwΔN MARCM clones ( hiwΔN+OE Kn ) rescued dendritic defects from 45% of reduction to 25% in dendritic length , and from 44% of reduction to 29% in dendrite termini number , as compared to wild-type ( Figures 5C–D ) , suggesting that Kn acts downstream of Hiw to control dendrite growth . In contrast , Kn overexpression had no effect on axonal growth in either wild-type or hiw mutant MARCM clones ( Figure 5E–F ) . Taken together , our results suggest that the Hiw-Wnd pathway acts through Kn to regulate dendritic , but not axonal , growth . There are four classes of dendritic arborization ( da ) neurons in Drosophila larva , which are categorized based on the complexity of dendritic branching [49] . Hiw mutations elevated the expression of puc-lacZ [50] , a reporter for Wnd activity [26] , in all four classes ( Figure S6A and B ) , suggesting that the Hiw-Wnd pathway is functional in all these neurons . However , Kn is only expressed in the class IV , and undetectable in other classes of da neurons [46]–[48] . If hiw acted via Kn to control dendritic growth , hiw mutations would not alter the dendritic morphology in class I ( C1 ) , class II ( C2 ) , and class III ( C3 ) da neurons . Indeed , we observed that hiw mutant MARCM clones of C1–C3 da neurons all exhibited normal dendritic growth ( Figures S7C and D , S8C and D , S9C and D ) , even though Hiw still restricts axonal growth in these neurons ( Figures S7A and B , S8A and B , S9A and B ) . These observations further suggest that the Hiw-Wnd pathway regulates dendritic growth in Kn-expressing neurons . We next determined whether Kn expression endows neurons with the ability to respond to dendritic growth control by Wnd . Consistent with previous reports that ectopic expression of Kn in class I da ( C1da ) neurons leads to excessive dendritic branching and extension [47] , [48] , the total dendrite length was increased by 55% and the number of dendritic branches was doubled in the C1da neurons overexpressing Kn ( OE Kn ) compared to wild-type . Such dendritic overgrowth was considerably reduced when Wnd was overexpressed in the same neurons ( Figure 6A–B ) , with the increase in total dendrite length inhibited from 55% to 10% . As a control , a kinase-dead form of Wnd failed to suppress Kn-induced dendritic overgrowth . Similar to the effects in C4da neurons ( Figure 4A–B ) , we detected a reduction of the nuclear Kn levels in C1da neurons expressing both Kn and Wnd ( Figure 6C–D ) . It is noteworthy that , in these C1da neurons , Kn was expressed by the Gal4/UAS system , which bypasses endogenous transcriptional control . Thus , up-regulated Wnd kinase is likely to suppress Kn expression via posttranscriptional mechanism . Collectively , these results suggest that Hiw-Wnd pathway regulates dendritic growth in Kn-expressing neurons by controlling the expression of Kn .
Taking into account the current study with previous studies , three distinct modes of axonal and dendritic growth regulation have been identified: shared , dedicated , and bimodal ( Figure 7A ) . Shared mechanisms co-promote or co-inhibit the growth of axons and dendrites . Molecular controls that operate in shared mechanisms include cytoskeleton regulators like MAP1B ( Futsch ) [51] , histone deacetylase HDAC6 [52] , [53] , and β-hexosaminidase [54] . Dedicated mechanisms provide the basis for specifically regulating the morphogenesis of only axons or only dendrites . Molecular controls at work in dedicated mechanisms can be divided into ( 1 ) axon-dedicated mechanisms , including p300 and SnoN transcription complex [2]; and ( 2 ) dendrite-dedicated mechanisms , including transcriptional factors NeuroD [3] and Dar1 [6] , growth factor BMP7 [4] , [5] , and small GTPase Rab17 [55] . Manipulation of dedicated mechanisms leads to specific changes in the growth of either axons or dendrites , but not both . Thus , axonal growth per se does not regulate dendritic growth , and vice versa . In contrast to dedicated mechanisms , bimodal mechanisms oppositely regulate axons and dendrites , and may serve to coordinate the growth of these separate compartments . Previous studies of different types of neuronal cultures have discovered three bimodal regulators: Sema3A [8] , [9] , CLASP2 [10] , and Rit [11] . In this study we have identified an in vivo bimodal regulatory mechanism that involves DLK kinase . The bimodal action of the DLK signaling pathway is achieved through two “dedicated” transcriptional programs . These two programs are likely to be independent because manipulating their activities rescues either dendritic or axonal defects , but not both , in hiw mutants . We also observed that transgenic Hiw and Wnd were present in the axon terminals in addition to the cell body but not in dendrites ( Figure S6C and D ) , raising the intriguing possibility that elevated Wnd function in the axon terminals might impact transcriptional activities in the cell body , and consequently influence denritic growth . It is likely that various bimodal controls exist in different neuron types . Moreover , it is possible that these bimodal controls intersect with each other . For instance , since the actions of Sema3A are mediated through cGMP/cAMP levels [9] , another bimodal regulator might also influence cGMP/cAMP levels . It will be interesting to determine whether cGMP/cAMP are involved in PHR-DLK pathway for bimodal control of dendritic and axonal growth . Despite the requirement of DLK functions in axonal growth after axon injury [24]–[28] , DLK is dispensable for axonal growth during development in the neuron types examined so far [14] , [17] . Consistently , we find that loss of dlk/wnd does not alter either dendritic or axonal growth in Drosophila C4da neurons . Rather , the overabundance of DLK/Wnd caused by defective PHR/Hiw functions leads to axonal overgrowth as well as dendritic reduction . Since axon injury leads to an overabundance of DLK/Wnd function [26] , [28] , it is conceivable that the elevated activity of DLK/Wnd induced by axon injury not only promotes axon regeneration [24]–[28] but also restrains dendritic growth or prunes exiting dendritic branches to compensate for the increased demand of membrane or cytoskeleton supplies for axonal growth . This notion is consistent with previous studies that show dendrite retraction following axotomy in Drosophila da neurons [56] and mammalian cultured neurons [57] , [58] . Although it is known that the zinc finger transcription factor Kn is essential for dendritic growth , the signaling mechanism that regulates Kn in neurons is unknown . In this study , we show that Kn specifically mediates dendritic regulation by the PHR-DLK pathway , which is supported by three lines of evidence . First , kn genetically interacts with hiw and functions downstream of hiw and wnd to regulate dendritic growth . Second , the Hiw-Wnd pathway regulates Kn expression in C4da neurons . Third , the Kn expression pattern is consistent with the presence of the Hiw-Wnd regulation of dendrite growth . Kn is selectively expressed in a subset of neurons [46]–[48] , [59] . Consistent with Kn expression pattern , hiw mutations caused dendrite defects only in the Kn-expressing class IV neurons , and not in the other classes of da neurons that lack Kn . Interestingly , ectopic expression of Kn in class I neurons , which do not normally express Kn , is sufficient to endow the Hiw-Wnd regulation . These results strongly suggest that the PHR-DLK pathway regulates Kn to control dendrite development . In contrast to Kn , the transcription factor Fos specifically mediates axonal regulation through Hiw-Wnd pathway . We found a two-fold role for fos in neuronal development . On the one hand , eliminating fos specifically causes dendritic reduction without affecting axon terminal length in C4da neurons . This indicates that endogenous Fos is specifically required for dendritic growth during normal development . On the other hand , the requirement of fos could switch to be axonal when augmented Wnd activity leads to exuberant axonal growth . In summary , the Hiw-Wnd pathway can exert bimodal or dedicated control over dendritic and axonal growth , depending on the presence of the transcription factors that mediate its subcellular compartment-specific functions . If transcription factors for both dendritic and axonal growth are present , Hiw-Wnd signaling functions as a bimodal modulator ( Figure 7B ) . This model provides guidance for further investigation of the molecular basis of the diversity of neuronal morphology and the differential development of dendrites and axons .
Fly stocks include hiwΔN [16]; hiwND8 [16]; UAS-Hiw::GFP [16]; wnd1 [17]; wnd3 [17]; UAS-Wnd [17]; UAS-WndK188A [17]; UAS-WndKD::GFP [17]; kay1 [60]; ca1 , FRT82B [61]; UAS-ltd::YFP [62]; kn1 [63]; knKN4 [63]; UAS-kn [46] , [64]; ppk-eGFP [34]; ppk-CD4::tdTomato [36]; ppk-CD4::tdGFP [36]; ppk-Gal4 [65]; UAS-Kinesin::βGal [37]; puc-lacZ [50]; UAS-RedStinger [66] . The MACRM analyses were performed as previously described [6] . For MARCM analyses of hiw mutations in four classes of da neurons , the tubP-Gal80 , hs-flp , FRT19A; Gal421-7 , UAS-mCD8::GFP virgins were mated with males of hiwΔN , FRT19A . For MARCM analyses of kay1 mutant , overexpressing Wnd , and overexpressing Wnd in kay1 mutant C4da neurons , the hs-flp; ppk-Gal4 , UAS-mCD8::GFP; FRT82B tubP-Gal80 virgins were mated with males of ( 1 ) UAS-Wnd; FRT82B , ( 2 ) FRT82B kay1 , and ( 3 ) UAS-Wnd; FRT82B kay1 , respectively . For MARCM analyses of ca1 mutations , the homozygous FRT82B ca1 virgins ( to remove maternal contribution of wild-type Claret ) were mated with males of hs-flp; ppk-Gal4 , UAS-mCD8::GFP; FRT82B tubP-Gal80 . To overexpress Kn in wild-type C4da neurons or in hiwΔN mutant C4da neurons , the tubP-Gal80 , hs-flp , FRT19A;; ppk-Gal4 , UAS-mCD8::GFP virgins were mated with males of FRT19A;;UAS-Kn and hiwΔN FRT19A;;UAS-Kn , respectively . Embryos and third instar larvae were dissected and immunostained as previously described [7] . The following primary antibodies were used: mouse anti-GFP ( Invitrogen , 1∶2 , 000 ) , chick anti-GFP ( 1∶2 , 000 ) , rabbit anti-RFP ( Rockland , 1∶2 , 000 ) , guinea pig anti-Knot ( gift from Adrian Moore , 1∶1 , 000 ) , rat anti-Elav ( DSHB , 1∶500 ) , guinea pig anti-Dar1 ( 1∶1 , 000 ) [6] , rabbit anti-Cut ( 1∶1 , 000 ) [67] , rabbit anti-βGAL ( Cappel , 1∶5 , 000 ) , and mouse anti-βGAL ( DSHB , 1∶100 ) . Confocal imaging was performed with a Leica SP5 confocal system . Only da neurons from abdominal segment 4 to 6 were imaged for quantification of dendrites and axons to ensure consistency . To compare protein expression levels in C4da neurons , larvae of different genotypes in the same experimental group were processed simultaneously . The same setting for image acquisition was applied to the same experimental group and signal saturation was minimized . Fluorescence intensities of different genotypes were normalized to wild-type ( Figures 4 and S5 ) or the OE WndKD control group ( Figure 6 ) . To quantify protein levels , mean fluorescence intensity of the region of interest in each channel was measured with NIH ImageJ software . For axon terminal and dendritic morphology , manual tracing was conducted with Neurolucida software . Branches shorter than 5 µm were excluded . For consistency , da neurons located between segment A4 and A6 from size-matched third instar larvae were imaged and analyzed in all experiments . In all of the bar charts of quantification , the numbers in the bars indicate the sample numbers . Values and error bars indicate mean ± SEM . Two-tailed unpaired student t-test was used . p values were indicated as: not significant ( NS ) p>0 . 05 , * p<0 . 05 , ** p<0 . 01 , *** p<0 . 001 . | Dendrites and axons are the input and output compartments of a neuron , respectively . Understanding how dendrites and axons are separated during neuronal development may help in developing strategies to correct defective neurons in neurological disorders and injuries . We show here that an evolutionarily conserved molecular pathway dichotomously controls dendritic and axonal growth . A key molecule in this pathway , dual leucine zipper kinase ( DLK ) , suppresses dendritic growth but promotes axonal growth . While DLK is known to be a key regulator of axon growth and regeneration , this study reveals its roles in dendritic growth for the first time . In addition , we find that the DLK pathway diverges through two separate downstream programs that control the expression of other genes . These insights can help target this pathway to specifically promote axon regeneration without affecting dendritic growth . Overall , these results help provide a new framework for understanding neuronal compartmentalization and the diversity of neuronal morphology . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"developmental",
"neuroscience",
"cellular",
"neuroscience",
"neuronal",
"morphology",
"biology",
"neuroscience",
"neural",
"circuit",
"formation"
]
| 2013 | Bimodal Control of Dendritic and Axonal Growth by the Dual Leucine Zipper Kinase Pathway |
Although flea-borne rickettsiosis is endemic in Los Angeles County , outbreaks are rare . In the spring of 2015 three human cases of flea-borne rickettsiosis among residents of a mobile home community ( MHC ) prompted an investigation . Fleas were ubiquitous in common areas due to presence of flea-infested opossums and overabundant outdoor cats and dogs . The MHC was summarily abated in June 2015 , and within five months , flea control and removal of animals significantly reduced the flea population . Two additional epidemiologically-linked human cases of flea-borne rickettsiosis detected at the MHC were suspected to have occurred before control efforts began . Molecular testing of 106 individual and 85 pooled cat fleas , blood and ear tissue samples from three opossums and thirteen feral cats using PCR amplification and DNA sequencing detected rickettsial DNA in 18 . 8% of the fleas . Seventeen percent of these cat fleas tested positive for R . felis-specific DNA compared to under two ( <2 ) percent for Candidatus R . senegalensis-specific DNA . In addition , serological testing of 13 cats using a group-specific IgG-ELISA detected antibodies against typhus group rickettsiae and spotted fever group rickettsiae in six ( 46 . 2% ) and one ( 7 . 7% ) cat , respectively . These results indicate that cats and their fleas may have played an active role in the epidemiology of the typhus group and/or spotted fever group rickettsial disease ( s ) in this outbreak .
Acute febrile flea-borne rickettsial diseases are caused by intracellular gram-negative bacteria Rickettsia typhi and Rickettsia felis , which are known to be transmitted to humans by the bite of Xenopsylla cheopis ( oriental rat flea ) and Ctenocephalides felis ( cat flea ) ; both fleas are found on many domestic and peri-domestic vertebrate hosts [1] . In recent years , additional flea transmitted Rickettsia has been identified in fleas and their hosts , both living in close proximity to people , warranting such rickettsial infections in humans referred to as flea-borne rickettsiosis . Unfortunately , when rickettsial infections are detected in humans the clinical diagnostic tests almost never distinguish between rickettsial species–R . typhi ( typhus group—TG ) or R . felis ( spotted fever group- SFG ) [2] . The commonly reported flea-borne rickettsiosis in humans is murine typhus whose known causative agent is R . typhi . Commercially available clinical serological diagnostic tests commonly show cross-reactivity among the pathogens in the TG and SFG . Confirmation of these infections requires positive serology in paired acute and convalescent serum samples of a four-fold or greater change in immunoglobulin G ( IgG ) and M ( IgM ) -specific antibody titer reactive to R . typhi or other Rickettsia species antigen by indirect immunofluorescence assay ( IFA ) . In most clinical settings , rarely are paired acute and convalescent serum samples collected , and California Department of Public Health resolved this shortcoming through special guidelines that achieved confirmation of murine typhus serologically or by nucleic amplification in a single serum specimen of elevated IgG and IgM antibody reactive to R . typhi or other Rickettsia species by IFA or DNA amplification , respectively ( CDPH 2011 ) [3] . Consequently , laboratory confirmation of murine typhus may not represent true cases , but cases of flea-borne rickettsiosis . Although these cases are reported as murine typhus without identifying the actual rickettsial pathogen but for purposes of consistent reporting they are reported as murine typhus . Here we identify such cases as flea-borne rickettsiosis and attempt to determine the responsible rickettsial agent for the current outbreak . On average , there are more than 200 human cases of murine typhus reported in the United States each year . This disease is not nationally reportable , so the true number of cases is unknown [4] . In areas where this disease is endemic ( California , Hawai’i , and Texas ) , providers and clinical laboratories are mandated to report cases to their local public health departments [2 , 5] . In California , most of the flea-borne rickettsial disease cases occur in Los Angeles and Orange counties , but concentrated outbreaks of the disease are rare [2] . Prior to 2015 , the last documented cluster of human flea-borne rickettsial disease in Los Angeles County occurred in 2009 [6 , 7] . Over the past decade , the incidence of flea-borne rickettsial disease in Los Angeles County ( LAC ) has increased . In 2014 , 51 cases were reported statewide [44 ( 86% ) in LAC] , 88 cases [69 ( 78% ) in LAC] in 2015 , and 90 cases [70 ( 78% ) in LAC] in 2016 . These cases in LAC occur in suburban communities through interactions between wildlife , domestic animals , and humans . Since the initial discovery of R . felis in C . felis in 1990 [8 , 9 , 10] , the association of R . felis with C . felis has been well-documented and the connection between zoonotic diseases and free-roaming animals has become an issue [8 , 9 , 10 , 11 , 7] . The last two decades have witnessed an increase in the recognition of new R . felis-like organisms ( RFLOs ) , particularly Rickettsia asembonensis and Candidatus Rickettsia senegalensis , whose distributions and host ranges appear to mimic those of R . felis [12] . Although there are no studies published concerning its ability to cause clinical illness , R . asembonensis was recently detected in the blood of monkeys ( Macaca fascicularis ) in Malaysia [13] , and in dogs’ blood in South Africa [14] . Another agent with close genetic composition to Ca . R . senegalensis was detected in the blood of febrile patients from Senegal [15] . Free-roaming animals have increasingly become a source of flea-borne infectious diseases that are controlled in domestic cat populations through routine veterinary care and flea control . When this care is lacking , the consequence is increased potential health risks for other domestic animals and humans [11] . Incidental infection transfer also occurs when free-roaming animals are fed outdoors . Wildlife , free-roaming cats , and domestic animals are brought in proximity when food is left outdoors , increasing the potential for exchanging fleas . When ectoparasites of wildlife become too numerous , they infest new hosts , increasing the risk that they may transmit disease to domestic animals and humans [16 , 17] . Three human case reports of flea-borne rickettsiosis among residents of a single 95-unit mobile home community ( MHC ) , with disease onsets from April 23rd to June 9th , 2015 prompted an initial investigation , and subsequent abatement order by the San Gabriel Valley Mosquito and Vector Control District ( District ) [18] . The Los Angeles County Department of Public Health ( LACDPH ) initiated active case finding to look for additional outbreak-associated cases during this time period and helped to coordinate a multi-agency notification , and abatement in the MHC [19 , 4] . The District conducted the abatement order in the MHC and coordinated overall surveillance , control , and mitigation efforts [20] . Here we discuss the outbreak , samples collected to identify the etiological agent ( s ) , and efforts made to mitigate the risk factors to public health .
In April 2015 , case A from the MHC was hospitalized for four nights and case B was hospitalized for six nights ( Table 1 ) . In June , case C from the same MHC was hospitalized for five nights [4] . A hospital infection preventionist reported the cases to LACDPH and they were forwarded to LACDPH Environmental Health Department and the District . From April to June 2015 , LACDPH conducted enhanced case finding for Rickettsia typhi-positive ( Quest Diagnostic , Inc . ; IFA test ) cases at all acute care facilities and local clinics within the catchment area of the MHC . Outbreak-associated cases were defined as MHC residents with symptom onset of fever with headache or rash from March 1st , 2015 through August 31st 2015 , and/or positive R . typhi or R . rickettsii laboratory test ( immunoglobulin M ( IgM ) >1:128 and/or immunoglobulin G ( IgG ) >1:128 . Additional criteria include elevated liver function tests ( ALT or AST ) , decreased platelet counts , and proximity in time and space for epidemiological-linkage to the outbreak . Cases A , B , and C were initially tested by Quest Diagnostics , Inc ( San Juan Capistrano , CA ) and two additional cases were tested and confirmed via indirect fluorescent antibody assays ( IFA ) by LAC Public Health Laboratory ( part of LACDPH ) . Paired acute and convalescent specimens of the cases were not available . Unlike LACDPH Laboratory , it’s not common practice for clinical commercial tests to include a full rickettsial panel of R . typhi and R . rickettsii . The State of California Department of Public Health guidelines interprets the detection of elevated IgG and IgM antibody reactive to R . typhi or other Rickettsia species antigen by IFA titer of ≥ 128 in a single serum specimen in addition to having specified clinical symptoms as confirmed cases [3] . These cases were part of anonymized LA County public health surveillance data . The initial investigation began in June 2015 with a survey of the property which included counting and photographing the number of pets and animals outdoors , the number of outdoor feeding sources ( water bowls and food ) , the presence and locations of free roaming animals , available harborage , and the presence of uncovered garbage bins . A door to door inspection and discussion with the residents of this MHC was conducted to inform them of the outbreak and symptoms of rickettsial disease , and the association of fleas and illnesses in the community . Discussions regarding appropriate cause of action were conducted between LACDPH , Environmental Health , and District staff . Opossums in the neighborhood were trapped using two live traps ( Tomahawk Live Trap , Tomahawk , WI ) set on selected properties in the MHC . Fleas were combed off the opossums trapped , identified to species level [21] , and tested by species-specific molecular methods for presence of rickettsial pathogens . A summary abatement order was issued by the District to the MHC on June 24 , 2015 to clean up the property . Specific requests included removal of animal feces , enforcement of MHC property rules limiting the pets to one animal per property , and regular flea control to reduce the population of fleas . The order also mandated that MHC residents cease outdoor feeding of pets and contract a pest control company to reduce the number of feral animals on the property . In addition , the property manager notified residents to provide flea control for all animals under their care , those tenants with more than one pet against the MHC homeowner rules and regulations to give them up and register their “single animal pet” with the property manager . This directive was necessary because most tenants had more pet animals some up to 32 , contrary to their property lease contracts . On August 24 , 2015 a multi-agency community event was held at a shopping center adjacent to the MHC by LACDPH , the District , the state Senator’s office District 20 , the city , and other regulatory agencies , to discuss public health risk posed by flea-borne rickettsiosis and encourage residents of the MHC to participate in reducing this risk . Residents of the MHC with symptoms of flea-borne rickettsial disease per case definition within the past three months were encouraged to provide blood samples for testing at no cost to them . Five individuals participated in the blood draw event and their blood tested by commercially available IFA tests , two of them were positive and was considered epidemiologically-linked human cases from this MHC . The tests were performed by Quest Diagnostics and/or confirmed by LAC Public Health Laboratory using commercially available FDA-approved IFA clinical tests . The property owner contracted with two pest control companies , one to control fleas and the other to remove feral animals on the property . The numbers of outdoor feeding sources were recorded monthly to determine the effectiveness of the property owners’ efforts . Flea control was conducted every 14 days , and the population of fleas was monitored bi-weekly by placing six 16 cm x 11 cm glue boards ( PIC Corporation , Linden , NJ ) throughout the property . Fleas collected on glue boards were counted and averaged by month to assess flea control efforts . Free-roaming domestic animals were counted monthly during morning walks of the property to monitor the impact of vertebrate/animal trapping efforts . A regression analysis ( JMP v 10/0: http://www . jmp . com ) was used to calculate the coefficient of determination ( static ) which correlated control measures conducted at the MHC against the number of outdoor wildlife , feeding sources , and flea activity . Fleas , blood , and ear tissue samples were retrieved from all cats removed by the vertebrate trapper and the opossums by the District from the MHC ( 13 cats , and 3 opossums ) for epidemiologic studies . One rat retrieved from the trap was dead and no flea , blood , or ear tissue samples were collected for testing . The Naval Medical Research Center tested all samples collected from the MHC epidemiologic study to determine the prevalence of rickettsial agents responsible for the outbreak . Fleas combed off animals trapped at the MHC were washed in molecular grade water and mechanically disrupted with disposable pellet pestles ( Fisher Scientific , Pittsburgh , PA ) . Genomic DNA was extracted with Prepman Ultra sample preparation kits ( Applied Biosystems , Foster City , CA ) . Genomic DNA from cat and opossum ear tissues and blood clots were extracted using the DNeasy blood and tissue kit ( QIAGEN , Valencia , CA ) according to the manufacturer’s instructions with a final elution volume of 50 μl . All fleas from cats ( n = 46 ) and 20 fleas from each opossum ( n = 60 ) were processed individually . Remaining fleas from the three opossums were processed in pools of 18–20 fleas ( n = 1 , 553 ) . Flea DNA , and DNA from the cat and opossum blood and tissues were initially screened for rickettsial DNA using a genus-specific quantitative real-time PCR ( qPCR ) assay ( Rick17b ) targeting the 17-kDa antigen gene [22] . DNA from the individual flea , cat , and opossum blood , and tissue samples that tested positive by the Rick17b assay were subsequently tested using a group-specific qPCR assay ( RfelB ) [23] and three species-specific assays , namely , ( 1 ) R . felis specific assay ( Rfel_phosp_MB ) that targets the membrane phosphatase gene from R . felis [24] , ( 2 ) the R . typhi species specific qPCR assay ( Rtyph ) which targets a fragment of the R . typhi ompB gene [23] , and ( 3 ) the Rickettsia asembonensis-specific qPCR assay ( Rasem ) , which targets a fragment of R . asembonensis ompB gene [25 , 26] . Similarly , the DNA from the pooled fleas was screened using the Rick17b , Rtyph , and the Rasem qPCR assays . PCR amplification and sequencing of gltA was attempted for a subset of 7 individual flea DNA including one DNA sample that was positive for rickettsial DNA using the Rick17b qPCR assay but negative for R . felis DNA based on the Rfel_phosp_MB qPCR assay , 2 samples that had discordant cycle threshold ( Ct ) values between Rick17b qPCR and Rfel phosp_MB qPCR assays and 4 flea DNA samples positive with both the Rick17b and the Rfel_phosp_MB qPCR assays to confirm the specificity of Rfel_phosp_MB . PCR amplification of gltA gene was attempted for all ear tissue DNA that tested positive for rickettsia DNA by qPCR , as previously described [20] . Sequencing reactions were performed in the forward and reverse directions utilizing the Big Dye Terminator v3 . 1 Reaction Cycle sequencing kit ( Life Technologies , Carlsbad , CA ) according to the manufacturer’s instructions using an ABI 3500 genetic analyzer ( Applied Biosystems ) . Sequence assembly was performed using CodonCode Aligner version 5 . 0 . 1 ( CodonCode Corporation , Centerville , MA ) . Blast searches were performed in NCBI websites . Evidence of previous infection of animals with spotted fever group ( SFGR ) and typhus group ( TGR ) rickettsiae was assessed using group-specific immunoglobulin G ( IgG ) enzyme-linked immunosorbent assays ( ELISAs ) as previously described [25 , 27 , 28] , using R . typhi str . Wilmington and R . conorii str . Morocco as the TGR and SFGR ELISA antigens , respectively . Serum samples were diluted 1:100 and screened for antibodies against rickettsiae using both the SFGR and TGR ELISAs . Screen positive samples ( those with net absorbance of ≥ 0 . 5 ) were titered by 4-fold serial dilution ( 100–6 , 400 ) . Commercial anti-cat IgG ( KPL , Gaithersburg , MD ) and anti-opossum IgG ( Alpha Diagnostics Intl . Woodlake Center , San Antonio , TX ) antibodies labeled with horseradish peroxidase ( HRP ) were used in both ELISAs . Ethics Statement: The San Gabriel Valley Mosquito and Vector Control District ( SGVMVCD ) do not have a formal Institutional Animal Care and Use Committee ( IACUC ) since it is not considered a research institution . However , it does follow the protocols for animal handling for disease surveillance purposes as outlined by the California Department of Public Health , Vector Borne Disease Section , and adhered to American Veterinary Medical Association ( 2013 ) guidelines for animal euthanasia . SGVMVCD as a cooperative member of Mosquito and Vector Control Association of California is exempt from requirement of holding a scientific permit under FG code 1002 , 4005 , and 4011 of the California Department of Fish and Wildlife ( CDFW ) to collect and sample small mammals for disease surveillance purposes .
Initially , three human ( A , B , and C ) cases were identified from the MHC with illness onset ranging from April to June 2015 . The LACDPH active case finding reported two additional epidemiologically-linked ( D and E ) cases of the outbreak for the period from March 1st through August 31st 2015 ( Table 1 ) . They were predominantly female ( 4/5 ) their ages ranged from 42 to 67 years; all were dog owners and two also owned cats . All experienced headaches , three had fever ( A , B , and C ) , and two ( D and E ) had a rash . The first three ( A , B and C ) cases were hospitalized for a total of 15 days ( mean 5 ) ; the remaining two cases were discovered from on-site blood draw at the community event ( D and E ) . The same three cases ( A , B , and C ) showed lower platelet count and elevated liver function enzymes ( ALT and AST ) . All cases recovered without complication . All five cases had antibodies reactive to R . typhi or other Rickettsia antigens with titers ≥ 1:128 via IFA testing except case B which together with cases C and D are considered epidemiologically-linked and part of the outbreak ( Table 1 ) . Two live traps set in June 2015 on the selected plots at the MHC yielded two opossums infested with 615 , and 1 , 087 cat fleas , respectively . Another opossum was trapped in September and was infested with 487 cat fleas . Flea control was conducted at the MHC every 14 days beginning in September 2015 by a professional pest control company . To assess its effectiveness , glue boards strategically distributed in the MHC at six locations were retrieved every two weeks from September to November . The glue boards placed on location prior to treatment contained as many as 39 fleas . The number of fleas on the glue boards declined significantly after the first treatment ( n = 36 , r² = 0 . 5813 , p≤ 0 . 05 ) , and the trend continued over time ( number of months ) , until no fleas were collected for two consecutive collection sessions ( Fig 1 ) . The MHC hired a wildlife trapper to remove outdoor vertebrate pests from the MHC . Thirteen cats were removed by the trapper from September to November . The number of fleas on the cats removed from the MHC declined over time ( n = 13 , r² = 0 . 7512 , p < 0 . 05 ) with the first cat collected in September having 13 fleas on it , and the one collected in November having no fleas . Subsequently , the number of cats outdoors from June through November also significantly declined from 29 to 4 ( n = 87 , r² = 0 . 914 , p < 0 . 05 ) over time ( Fig 2 ) . The number of outdoor feeding stations at the MHC also decreased significantly ( n = 90 , r² = 0 . 9697 , p < 0 . 05; Fig 2 ) . The decline in the population of fleas on the ground and on cats , free-roaming animals , and wildlife , and the decreased prevalence of outdoor feeding ultimately decreased the risk at the MHC of acquiring flea-borne typhus . A total of 13 cats and three opossums were trapped around the MHC over the study period . Pooled fleas ( n = 1 , 553 ) combed from these animals were screened for rickettsiae with the genus specific qPCR assay ( Rick17b ) , R . typhi specific qPCR assay , and R . asembonensis- specific qPCR assay . Eighty-five pools were tested , with 74 ( 85 . 33% ) testing positive for rickettsiae ( Table 2 ) . Further screening was qPCR assay negative for both R . typhi and R . asembonensis . One hundred and six individual fleas from cats and opossums were tested . From 9 of 13 cats , 46 fleas , and from three opossums , 60 fleas making a total of 106 fleas were tested for rickettsial DNA ( Table 3 ) . Twenty of these fleas ( 18 . 8% ) were positive for Rickettsia DNA using the genus-specific assay Rick17b . Of these , 18 were positive for R . felis using the Rfel_phosp_MB assay . None of the fleas were positive for R . typhi or R . asembonensis-specific qPCR assays . To confirm the identity of rickettsiae identified by the qPCR assays , PCR amplification and sequencing of the gltA gene was done in a subset of seven individual flea DNA preparations . Seven 1186-bp gltA sequences were generated , of which five were 100% identical to each other and to R . felis URRWXCal2 ( Accession no . CP000053 ) . The five samples included four that had tested positive for both Rick17b and Rfel_phosp_MB , and one of the two samples that had discordant Ct . The remaining two sequences were 100% identical to each other and to Candidatus Rickettsia senegalensis ( Accession number KF666472 ) , and included one that was positive for Rick17b but negative for Rfel_phosp_MB , assay , and one that had discordant Ct values . A total of 13 cat sera and three opossum sera were assessed for antibodies against TGR and SFGR by ELISA and the presence of Rickettsia DNA by qPCR assays . Of the sera tested for the presence of SFGR- and TGR-specific IgG , 6/13 ( 46 . 15% ) had IgG antibodies reactive against TGR antigens with endpoint titers ranging from 1600 to 6400 ( Table 4 ) . One of the 13 cat sera ( 7 . 69% ) was positive for antibodies against SFGR antigens with an endpoint titer of 100 . None of the opossum sera were positive for antibodies against SFGR- or TGR-specific IgG . The blood clots and tissues from 13 cats and 3 opossums were also assessed for the presence of Rickettsia DNA . All blood clot DNA preparations from cats and opossums were negative for Rickettsia DNA whereas three of thirteen cat tissues tested positive for Rickettsia DNA ( Rick 17b with Ct values > 35 ) . One of three cat tissue DNA preparations tested positive for R . felis by group-specific qPCR RfelB assay , but all three cat tissue preparations were negative for other PCR assays and none produced amplicons for sequencing .
An outbreak of five cases of flea-borne rickettsial disease occurred at the MHC within the District . It is highly likely that additional infections occurred due to the abundance of fleas on the property , but went undetected due to a mild presentation and/or lack of testing . The etiologic agent of flea-borne rickettsial diseases has been debated for years . R . typhi has been known historically as the etiologic agent of murine typhus [29 , 1 , 30 , 2] . There were three hospitalized and two non-hospitalized epidemiologically-linked cases at the MHC that had R . typhi or other Rickettsia antibody titers ≥ 1:128 except one of hospitalized cases with titers ≥ 1:64 via IFA test , all in private clinical laboratory or/and LAC Public Health Laboratory . Although human testing met state guidelines the only limitation of the clinical diagnostic testing of all five cases was the use of a single serological sample instead of the paired acute and convalescent samples . Such samples were not available because it is uncommon for clinicians to collect paired samples from patients in clinical settings . Physicians often overlook rickettsial infections and by the time such tests are deemed necessary for patient care a dose of antibiotics/chemotherapy would have been administered rendering collection of paired serological sample impossible . The clinical IFA test conducted on single patient sera does not confirm murine typhus but confirms rickettsial infection either of TG or SFG . The epidemiology of rickettsial diseases in southern California , and especially San Gabriel Valley in Los Angeles County show that R . typhi is the predominant human Rickettsia pathogen compared to SFG-transmitted R . rickettsii , thus it has not been important from a public health standpoint to differentiate between TG and SGF infections [2 , 4] . Current assays examining IgG/IgM titers with indirect fluorescent antibody assays do not always differentiate between antibodies against R . typhi and other Rickettsia species , and the former is assumed to be the etiologic agent for murine typhus , therefore of flea-borne rickettsiosis [31] . A more accurate differentiation between R . typhi and other Rickettsia species could be made through PCR based assays if a blood specimen could have been available from acute symptomatic individuals . Unfortunately , at the time of investigation , acute blood specimens from the three hospitalized cases were unavailable . Samples from the animals and fleas removed from the MHC were tested for the presence of the different etiologic agents . Rickettsia felis was detected in fleas obtained from animals at the MHC , which supports prior research that elevated the potential role of R . felis previously referred to as ELB as an etiological agent of flea-borne rickettsial disease in relation to R . typhi , and implicated opossums and rodents as the main hosts [2 , 32 , 33] . However , results showed that both R . felis and Ca . R . senegalensis were present in cat fleas , although the association of R . felis with human disease seems poor [34] . The focus was on R . typhi where 46 . 15% of cat sera removed from the MHC were positive , and none of the opossum blood had any rickettsial DNA . Low level IgG-positive opossum and lack of rickettsial DNA in opossums aligns with findings from previous studies [12 , 34] . Alternatively , the presence of antibodies against R . typhi in cats suggests that it could still be active within the peri-domestic and domestic animal community and this may suggest that cats provide another mechanism for maintaining typhus in southern California . Furthermore , R . typhi was not detected in the cat fleas , which corroborates past findings [35] that R . typhi may still be present in the environment at a very low but infectious level and/or possibly carried by another flea species and hosted by other mammals beside opossum and cats . Beside R . felis , the present study detected Ca . R . senegalensis in cat fleas . This corroborates the findings of two previous studies that reported existence of the “RFLO” in southern California [12 , 35] . Although the previous studies reported R . asembonensis at a relatively low rate ( 0 . 3% of 597 fleas tested ) in the same region [12] , the present study did not confirm that finding . Although two RFLOs have been detected in the blood of dogs , monkeys , and humans [15 , 13 , 14] , their ability to cause disease in mammalian species has not yet been proven . The relationship ( symbiosis , mutualism , or parasitism ) between the host and the rickettsiae has not been elucidated for any of the RFLOs . It has been suggested that all rickettsiae can potentially be pathogenic to vertebrate hosts [36] . This is evidenced by the findings that R . slovaca , R . helvetica and R . parkeri tick endosymbionts were associated with human disease years after they were discovered [37 , 38 , 39] . Within five months of focused abatement implementing mitigation measures , the potential risk of rickettsial and other flea-borne diseases infection was reduced based on several observations within the neighborhood under investigation . The achievement of mitigation measures involved flea control , trapping and removal of feral cats , and opossums , providing flea-collars to residents for cats and dogs , and removal of outdoor feeding sources by the property owner and tenants . The role of public health agencies was education and coordination , and that of vector control was identifying the responsible parties–property owner and tenants–and ensuring they contracted with professional pest control whose work was certified at completion . More importantly no additional cases of flea-borne rickettsioses were detected . Ultimately , the success of this approach as spearheaded by public health agencies was measured by the absence of new cases of flea-borne typhus , but the MHC must continue to adhere to its policies to ensure that public health risk previously present do not re-occur . | Outbreaks of flea-borne rickettsiosis are rare despite the endemic status in Los Angeles County . In the spring of 2015 three human cases of flea-borne rickettsiosis among residents of a mobile home community ( MHC ) prompted an investigation . Fleas were found in all common areas at the MHC due to presence of flea-infested opossums and overabundant outdoor cats and dogs . The MHC was summarily abated in June 2015 , and within five months , flea control and removal of animals significantly reduced the flea population . Two additional epidemiologically-linked human cases detected at the MHC were considered to have occurred before control efforts began . Molecular testing of cat fleas , immunological testing of opossums and feral cats collected at the site indicated active transmission of flea-borne rickettsiosis . This study represents the first flea-borne rickettsial outbreak that summary abatement approach was used to reduce its intensity . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
]
| [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"vertebrates",
"rickettsia",
"mammals",
"animals",
"marsupials",
"clinical",
"medicine",
"fleas",
"infectious",
"disease",
"control",
"bacteria",
"bacterial",
"pathogens",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"major",
"histocompatibility",
"complex",
"medical",
"microbiology",
"microbial",
"pathogens",
"insects",
"arthropoda",
"opossums",
"eukaryota",
"blood",
"anatomy",
"cats",
"clinical",
"immunology",
"physiology",
"biology",
"and",
"life",
"sciences",
"amniotes",
"organisms"
]
| 2018 | A 2015 outbreak of flea-borne rickettsiosis in San Gabriel Valley, Los Angeles County, California |
A major barrier to curing HIV-1 is the long-lived latent reservoir that supports re-emergence of HIV-1 upon treatment interruption . Targeting this reservoir will require mechanistic insights into the establishment and maintenance of HIV-1 latency . Whether T cell signaling at the time of HIV-1 infection influences productive replication or latency is not fully understood . We used a panel of chimeric antigen receptors ( CARs ) with different ligand binding affinities to induce a range of signaling strengths to model differential T cell receptor signaling at the time of HIV-1 infection . Stimulation of T cell lines or primary CD4+ T cells expressing chimeric antigen receptors supported HIV-1 infection regardless of affinity for ligand; however , only signaling by the highest affinity receptor facilitated HIV-1 expression . Activation of chimeric antigen receptors that had intermediate and low binding affinities did not support provirus transcription , suggesting that a minimal signal is required for optimal HIV-1 expression . In addition , strong signaling at the time of infection produced a latent population that was readily inducible , whereas latent cells generated in response to weaker signals were not easily reversed . Chromatin immunoprecipitation showed HIV-1 transcription was limited by transcriptional elongation and that robust signaling decreased the presence of negative elongation factor , a pausing factor , by more than 80% . These studies demonstrate that T cell signaling influences HIV-1 infection and the establishment of different subsets of latently infected cells , which may have implications for targeting the HIV-1 reservoir .
HIV-1 persists in a transcriptionally silent latent state in long-lived memory T cells . Although antiretroviral therapies ( ART ) suppress HIV-1 replication , interruption of treatment results in rapid viral rebound . Therefore , HIV-1 patients must remain on ART indefinitely , despite long-term side effects , development of treatment resistance , and viral-induced inflammation [1–3] . For this reason , one strategy currently being explored for cure efforts is “shock and kill , ” in which latent HIV-1 is reactivated in conjunction with ART using latency-reversing agents ( LRAs ) . Following reactivation , infected cells are predicted to be eliminated by HIV-specific immunity or virally induced apoptosis . However , clinical trials using LRAs have only minimally perturbed the size of the viral reservoir [4–6] . A cure for latent HIV-1 will require a better understanding of the biochemical factors involved in regulating proviral transcription . Latency in chronically infected primary cells and cell lines is regulated by multiple transcriptional mechanisms including NF-κB activation , chromatin accessibility , provirus transcription initiation , Tat availability , P-TEFb sequestration , and transcriptional elongation [7–11] . However , what is not understood is how latency is initially established within a cell and if events at the time of HIV-1 infection influence the transcriptional status of the provirus . These questions are relevant since the latent reservoir is established within the first two weeks of infection [12 , 13] . New cure strategies will need to limit the size of the reservoir at early time points . One mechanism that could predispose HIV-1 towards active replication or transcriptional repression and latency is signaling through the T cell receptor ( TCR ) . Engagement of the TCR and costimulatory CD28 molecule result in a multitude of cellular outcomes that influence HIV-1 replication including cytoskeleton reorganization , the activation of transcription factors , enhanced RNA polymerase II ( RNAP II ) processivity , and chromatin remodeling [14–16] . We hypothesized that the magnitude of T cell signaling during HIV-1 infection will dictate the course of the infection . In order to manipulate signal strength received by a T cell at the time of HIV-1 infection , we utilized chimeric antigen receptors ( CARs ) that recapitulate T cell receptor and CD28 signaling . By modulating the affinity with which these CARs bind to their ligand , we can differentially deliver signals to target cells . Using these CARs , we demonstrate that stronger T cell signaling at the time of HIV-1 infection increases subsequent HIV-1 transcription and replication . Robust signals also facilitated the formation of latently infected cells that were readily inducible upon secondary stimulation . Minimal signaling through CARs , although sufficient for HIV-1 integration , failed to support viral replication and generated a deep-seated latent infection . Transcriptional elongation of HIV-1 provirus was limited by RNAPII pausing in the absence of CAR signaling; however , strong CAR signaling correlated with decreased negative elongation factor ( NELF ) binding and enhanced RNAPII processivity . Our results suggest a model in which signaling strength influences HIV-1 transcription and establishment of latency at the time of initial infection of CD4+ T cells .
To examine how signaling cascades downstream from the T cell receptor regulate HIV-1 transcription we utilized CARs ( Fig 1A ) . Intracellular signaling domains for the CARs include CD3ζ with its immunoreceptor tyrosine-based activation motifs ( ITAMs ) and the CD28 costimulatory domain with its four critical tyrosine residues [17] . Furthermore , a mCherry tag provides a marker for positive selection of CAR+ cells . The extracellular ligand-binding domains of the CARs consist of a single chain variable fragment ( scFv ) that recognizes receptor tyrosine-protein kinase erbB-2 ( Her2 ) [18 , 19] . By using different scFvs , a library of CARs with binding affinities for Her2 ligand spanning three logs was generated ( Fig 1B ) . CARs were transduced into Jurkat T cells and primary CD4+ T cells . By enriching for mCherry , we obtained CAR+ populations that were >90% pure ( Fig 1C ) . We confirmed that signaling through CARs mimicked aspects of TCR signaling and supported differential changes in both downstream gene expression and T cell phenotypes . CD69 , a transmembrane lectin and a marker for CD4+ T cell activation , was monitored by flow cytometry before and after receptor activation with Her2 ligand ( Fig 2A ) . Primary CD4+ T cells transduced with either the low affinity or the high affinity receptors were stimulated with plate-bound Her2 ligand for 24 h . In the absence of ligand , less than 7% of the cells were positive for CD69 , verifying that there is no ectopic CAR signaling . Activating cells with Her2 induced CD69 expression in the low affinity and high affinity receptors relative to their affinity for ligand . In addition , we analyzed the ability of the receptors to generate T cell subsets one week post stimulation by determining the expression of CCR7 , a lymph node homing receptor , and CD45RA , a marker that is downregulated on memory subsets following activation through endogenous TCR . Similar expansion of CCR7+ CD45RA- T cell populations was observed in cells stimulated through the TCR and both the high and low affinity CARs ( Fig 2B ) . Finally , microarray analysis was performed to determine if CAR and TCR signaling resulted in similar global gene expression changes . A heatmap of genes whose expression was significantly altered compared to baseline upon activation through either the high and low affinity CARs or treatment with anti-CD3 and anti-CD28 is shown in S1 Fig . A spectrum of changes in gene expression downstream from the T cell receptor was observed that included subsets of genes that were induced or repressed to similar extents by the TCR and CARs as well as more graded differential responses in which the high and low affinity CARs induced intermediate changes as compared to the T cell receptor . We did observe donor-to-donor variation , which was expected . This may reflect intrinsic donor variation as well as sexual dimorphism because of the inclusion of 1 male and 2 female donors . Together with the CD69 analysis and memory cell markers , these data demonstrate that signaling through the chimeric antigen receptors facilitate a range of cellular effects associated with T cell activation and that the CARs can be used as tools to modulate T cell signals . To determine whether T cell signaling influences viral infection , Jurkat T cells expressing low affinity or high affinity CARs were plated on Her2-coated wells and simultaneously infected with VSV-G pseudotyped NL4-3 . Luc , a single-cycle HIV-1 clone which contains a luciferase reporter in place of Nef . VSV-G allowed us to bypass potentially confounding effects from receptor/chemokine receptor signaling due to gp120 binding and focus specifically on CAR-associated signaling cascades . To assess whether signaling influenced the establishment of infection , we measured levels of HIV-1 proviral DNA using a previously described nested Alu-PCR approach [20] . We modified the assay by designing primers to luciferase to estimate the relative frequency of HIV-1 integration without confounding signals from the lentiviral vectors used to express the CARs ( see Materials and Methods ) . CAR-associated signaling did not affect the infection of Jurkat cells since we detected comparable levels of provirus regardless of the presence or absence of CAR ligand ( Fig 3A ) . When HIV-1 expression was measured by luciferase activity , Jurkat cells infected in the context of strong T cell signaling expressed greater than 10-fold more HIV-1 compared to untreated controls ( Fig 3B ) . In contrast , engagement of the low affinity receptor led to a modest 3-fold expression compared to unstimulated cells despite a similar proviral load as the high affinity CAR-expressing cells . These data indicate that strong T cell signaling at the time of infection facilitates HIV-1 expression without enhancing provirus integration . We confirmed that these differences were due to downstream signaling emanating from the CARs by using the Src kinase inhibitor PP2 . In the presence of PP2 , the increase in HIV-1 expression upon cellular stimulation was attenuated , consistent with T cell signaling as a regulator of HIV-1 expression ( S2 Fig ) . The pharmacologically inactive version of this inhibitor , PP3 , had no effect on the ability of CARs to influence HIV-1 expression . We validated these results using primary CD4+ T cells that were transduced with either the low affinity or high affinity CAR . Following transduction , cells were allowed to return to a resting state as monitored by low CD69 expression before infection with HIV-1 in the absence or presence of the ligand Her2 . Consistent with the data from Jurkat cells , similar levels of proviral DNA were detected in primary T cells regardless of CAR signaling ( Fig 3C ) . Depending on the donor , cells that received robust stimulation at the time of infection expressed 2 . 5- to 11-fold more HIV-1 compared to untreated controls . Stimulating through the low-affinity CARs led to more modest luciferase expression when compared to untreated cells ( Fig 3D ) . To gain insight into whether there is a threshold or minimal T cell signal required for HIV-1 infection and replication , we transduced Jurkat T cells with CARs that spanned a range of binding affinities ( Fig 1B ) . These cells were infected with NL4-3 . Luc as described above in the absence or presence of Her2 . Although the high affinity condition supported HIV-1 infection and transcription , the intermediate and low affinity receptors did not support HIV-1 expression ( Fig 4A ) . This was despite similar levels of infection as determined by measuring proviral DNA ( Fig 4B ) and a greater than 10-fold increase in binding affinity for the Her2 ligand above that of the low affinity CARs . These data suggest that T cell signaling controls HIV-1 expression by a digital on/off mechanism since viral expression does not linearly correlate with signal strength . We hypothesized that differential T cell signaling during infection alters the size of the inducible latent reservoir . To examine this , we infected CAR-expressing primary CD4+ T cells with VSV-G pseudotyped BRU-dENV-GFP in the presence of Her2 ligand . One week post infection , cells were sorted for both mCherry expression as a marker for the CAR and lack of GFP expression in order to enrich for latently infected cells . CARpos/GFPneg cells were reactivated with PMA plus ionomycin or left unstimulated to control for spontaneous HIV-1 reactivation ( Fig 5A ) . PMA plus ionomycin significantly reactivated HIV-1 expression within cells that had been initially infected in the context of high-affinity CAR , resulting in a 3- to 9-fold increase in the percentage of GFP positive cells ( Fig 5B ) and a 1000-fold induction of HIV-1 mRNA measured by qRT-PCR ( Fig 5C , S3 Fig ) . However , the observed reactivation of HIV-1 was modest in cells infected at the time of stimulation through the low affinity CAR . Less than a 2-fold change was observed in the percentage of GFP+ cells , and only a 200-fold induction of HIV-1 mRNA was detected in reactivated latently infected cells expressing low affinity CARs . Thus , there was a ~5-fold increase in HIV-1 mRNA induction for cells expressing high affinity CAR upon reactivation compared to reactivation in cells expressing low affinity CAR ( Fig 5C , S3 Fig ) . A panel of latency reversing agents were tested for their abilities to reactivate the latently infected cells generated by the different CARs . Cells were reactivated with antibodies to CD3 and CD28 , the HDAC inhibitor SBHA , and the PKC agonist Bryostatin ( Fig 5B ) . Secondary stimulation through the endogenous T cell receptor with anti-CD3+CD28 reactivated latent HIV-1 in cells that had been infected and stimulated through the high-affinity receptor , resulting in a 3- to 7-fold increase in the percentage of GFP positive cells . However , anti-CD3+CD28 treatment resulted in either no reactivation or a modest 1 . 8-fold reactivation in cells stimulated through the low affinity CAR at the time of HIV-1 infection . SBHA did not strongly induce HIV-1 expression but there was a trend of greater virus reactivation in cells that had received stronger TCR signaling at the time of HIV-1 infection . Treatment with Bryostatin did not lead to robust activation for either low affinity or high affinity CAR-expressing cells . In general , there appeared to be different reservoir sensitivities to latency reversal agents between cells that had received strong or weak signaling during infection , especially when comparing cells derived from the same donor . Our data suggest that despite comparable amounts of integrated HIV-1 proviruses , robust signaling at the time of infection was not only necessary for active proviral transcription but also supported the generation of a population of latently infected cells that could be readily induced to express HIV-1 . The population of latent cells generated in response to weaker CAR signaling was more resistant to latency reversal suggesting that HIV-1 in these cells was strongly repressed . We were interested in mechanisms that governed HIV-1 repression following integration in the absence of sufficient T cell signaling; therefore , we examined the binding of transcriptional regulators on the HIV-1 LTR by chromatin immunoprecipitation ( ChIP ) . Jurkat T cells expressing low or high affinity CARs were infected with NL4-3 . Luc in the absence or presence of Her2 ligand . One day post-infection , cells were fixed and chromatin was prepared for ChIP . Since HIV-1 proviral latency correlates with a positioned nucleosome that is downstream of the transcriptional start site , we explored whether the LTR was associated with post-translationally modified histones as an indicator of chromatin organization . ChIPs for acetylated histone H3 showed no significant difference in binding of the HIV-1 LTR between cells infected in the absence or presence of T cell signaling ( Fig 6A ) . Therefore , chromatin accessibility does not appear to be limiting HIV-1 proviral transcription following infection . We then examined RNAP II processivity by measuring RNAP II occupancy at multiple points , including the transcriptional start site and downstream in the HIV-1 tat gene . RNAP II was detected at the HIV-1 transcriptional start site whether cells were activated through a CAR or were unstimulated ( Fig 6B ) . However , signaling through the high affinity receptor resulted in an increase in downstream RNAP II by greater than 4-fold , whereas only modest levels of RNAP II were found downstream in the absence of signals or following weak signaling ( Fig 6C ) . Since these data indicated a role for transcriptional pausing , we examined if CAR signaling altered the presence of the pausing factor negative elongation factor ( NELF ) at the HIV-1 transcriptional start site . Using ChIPs , we determined that signaling through the high affinity receptor diminished binding of NELF at the HIV-1 LTR by greater than 85% ( Fig 6D ) . These data support a model in which a lack of robust T cell signaling limits HIV-1 transcription by establishing a paused polymerase complex .
Previous studies suggest that cell signaling may be a key regulator of HIV-1 expression and latency . The latent reservoir is enriched for antigen specific T cells , including those that respond to CMV , HSV , tuberculosis , and HIV [21–25] . Furthermore , the use of superantigens during viral entry increases HIV-1 replication [26] . Partial activation , cellular polarization , cell-to-cell contact , and/or infection of resting quiescent cells through perturbation have also been suggested to bias infections towards latency [11 , 27–31] . Therefore , the extent of cell activation is a key determinant in regulating the course of HIV-1 infection including the formation of the reservoir . We have shown that differential signaling through CARs , which mimic TCR signaling , influences HIV-1 transcription and latency . In the lymph node , a primary site for both HIV-1 replication and the persistent latent reservoir [32–34] , T cells will sample lymph node resident cells in search for antigen . Some of these interactions , facilitated by the presentation of the T cell cognate antigen , will result in robust T cell activation , clonal expansion , and changes in gene expression . However , most MHC complexes will lack cognate antigen and initiate weak signaling [35 , 36] . Using multiple CARs whose affinities for the Her2 ligand span several logs , we can deliver a range of signaling inputs to model the spectrum of T cell receptor signaling events . Our data indicates that stronger T cell activation at the time of infection , which would be more similar to antigen specific responses , correlates with robust HIV-1 expression as well as the establishment of inducible latently infected cells . We validated these findings in both Jurkat cell lines and primary CD4+ T cells derived from multiple donors although the magnitude of responses from primary cells was more variable as would be predicted . Additional factors may compensate for suboptimal T cell receptor signaling including cytokine-induced stimulation and interactions with antigen presenting cells that would engage both costimulatory molecules and inhibitory receptors . The contribution of the T cell receptor pathway and how this is integrated with other signaling events to influence HIV-1 infection and latency is a critical question that needs to be addressed . Having a library of CARs with a range of binding affinities allowed us to determine if HIV-1 responds to signaling in an analog fashion correlating with signal input or is digitally regulated by specific thresholds resulting in all-or-none responses [37] . Signaling through the CARs with affinities that were intermediate did not support active transcription despite a greater than 10-fold increase in binding affinity compared to our low affinity receptor . These results would suggest that TCR signaling provides more of an on/off switch in regulating HIV-1 transcription and that there exist signaling thresholds that must be overcome to assure efficient HIV-1 transcription and replication . Signal transduction and gene expression are inherently noisy processes , and stochastic events are hypothesized to drive HIV-1 latency . That latency and HIV-1 replication are driven by episodic bursts of proviral transcription and Tat levels has been supported by mathematical modeling and experiments using engineered virus models [38–40] . Even if latency is driven by random fluctuations of provirus transcription , T cell associated signals are strong modulators of noise , and targeting these pathways could enhance treatments directed at HIV-1 reactivation [41] . Weak signaling , such as those induced by the low affinity chimeric antigen receptors , may be inadequate to alter the inherent noise within the system , whereas robust TCR signals through the high affinity CAR increase the probability of stochastic events . It is important to appreciate that although signaling and transcription are subject to stochastic variation , these are coordinated and combinatorial processes that lead to defined patterns of gene expression and phenotypic outcomes [42] . Regulated aspects of transcription include assembly of multi-subunit complexes such as RNAP II and associated cofactors , chromatin , and transcription factors at the LTR . Our data suggest that the association of NELF with RNAP II is regulated by TCR signaling . Multiple positive and negative signals are known to converge on NELF-driven transcriptional pausing . P-TEFb relieves NELF repression through phosphorylation [43] and is itself regulated by cellular stress and signals [44–46] . Furthermore , we have shown that NELF interacts with co-repressors including NCoR1-GPS2-HDAC3 at the HIV-1 promoter [47] which may reinforce HIV-1 latency , especially during chronic infection , by facilitating post-translational modifications of histones and chromatin organization . We propose that strength of signal at the time of infection acts as a bifurcating event leading to either robust transcription and the establishment of an inducible latent reservoir or minimal transcription and deep-seated latency . Our observations are consistent with the previous characterization of patient reservoirs that identified three subsets of latently infected cells: a small population of cells carrying inducible provirus , a larger population of cells with intact proviruses that are difficult to reactivate , and many defective proviruses [48] . Successful purging of the latent reservoir may require the use of a cocktail of latency reversing agents or the development of novel strategies to block reactivation [49–51] .
Jurkat CD4+ T cells ( E6-1 ) and human embryonic kidney 293T cells were obtained from American Type Culture Collection ( ATCC ) . Jurkat cells were cultured in RPMI 1640 , 5% FBS ( Corning , Inc . ) , 100 units/mL penicillin ( Invitrogen ) , 100 μg/mL streptomycin ( Invitrogen ) , and 2mM L-glutamine ( Invitrogen ) . HEK293T cells were cultured in Dulbecco’s Modified Eagle Medium , 10% FBS , 100 units/mL penicillin , 100 μg/mL streptomycin , and 2mM L-glutamine . Cells were grown at 37° C with 5% CO2 . Primary CD4+ T cells were derived from de-identified healthy blood leukapheresis packs purchased from NY Biologic . Mononuclear cells were enriched from leukopaks by centrifugating through Histopaque gradient ( Sigma-Aldrich ) . CD4+ T cells were isolated by negative selection using EasySep Human CD4+ T Cell Enrichment Kits from STEMCELL Technologies . CD4+ cells were maintained in RPMI 1640 , 10% FBS , 100 units/mL penicillin , 100 μg/mL streptomycin , and 2mM L-glutamine at 37° C with 5% CO2 . Prior to transduction with CARs , primary cells were supplemented with 10 units/mL IL-2 and 10 ng/mL IL-7 . Following transduction , IL-2 was removed from most culture conditions . All cells and cell lines were split every 2–3 days . CARs were driven by a SFFV promoter in the lentiviral vector pHR [18 , 19] . pNL4-3 . Luc . R-E- was obtained from NIH AIDS Reagent Program . BRU-ΔEnv-GFP has been described before [52] . Lentiviruses were made by transfection of vectors , VSV-G , Rev , Tat , and Gag-Pol constructs into HEK293T cells with 45μL polyethylenimine ( 1 mg/mL ) per 6x106 cells . Supernatants were collected , filtered with 0 . 45μm syringe filter ( Corning ) , concentrated by centrifuging through a 20% sucrose gradient , and titered with CEM cells [53] . We used a range of multiplicity of infections , but most viruses and lentiviruses within this paper were concentrated to approximately 1x106 IU/mL . HIV-1 viruses were made similarly but only required the viral plasmid and VSV-G . For transductions with CAR vectors , a minimum of 1x106 primary and Jurkat cells were stimulated for 5–6 h with 10 μg/mL PHA , washed in PBS , and spinoculated with lentivirus and 5 μg/mL polybrene ( Millipore ) at 1200g for 90 min . Cells were then supplemented with fresh RPMI and IL-7 , cultured overnight , and washed in PBS 18 h later . Cells were rested for one week to return to a resting state as confirmed by low CD69 expression prior to HIV-1 infection . Non-tissue culture treated plates were coated overnight at 37°C with 1 μg/mL Her2 ( Recombinant Human ErbB2/Her2 Fc Chimera Protein from R&D Systems , 1129-ER ) . Her2 solution was removed from wells , plates were washed 3 times in PBS , and wells were blocked for 1 h with a 5% FBS-PBS solution . Jurkat or primary CD4+ T cells were infected and simultaneously plated in Her2-treated wells . For experiments in which latently infected cells were generated , cells were spinoculated in the Her2-treated wells at 1200xg for 90 min and then supplemented with fresh RPMI and IL-7 . Following overnight infection , cells were washed and either lysed or maintained in fresh media in the absence of Her2 . For reactivation of latent cells , mCherry ( CAR ) positive and GFP ( HIV ) negative cells were sorted at 6 or 7 days post HIV-1 infection . Cells were cultured with the following concentrations of LRAs: 5 ng/mL PMA ( Fisher Scientific ) and either 10 or 100uM ionomycin ( Sigma-Aldrich ) for 2 . 5 h , Dynabeads human T-activator CD3/CD28 beads at a ratio to cells of 1:1 for 24 h , 50 μM SBHA ( Sigma-Aldrich ) for 24 h , and 25nM Bryostatin ( Sigma-Aldrich ) for 24 h . Cells reactivated with PMA were washed in PBS and re-plated in media . All reactivated cells were incubated with 10 ng/mL IL-7 . Cells were cultured overnight prior to fixation for flow analysis . For some experiments , cells were treated with 10 μM PP2 or PP3 ( Calbiochem—Millipore Sigma ) at the time of infection . Flow data were collected on an LSRII from BD Biosciences . Zombie UV Fixable Viability Kit ( BioLegend ) was used as live/dead stain for reactivation experiments . We had a range of 250 to 6000 events per data point and a minimum cut-off of 250 events in Fig 5B . The mean number of all events was 1364 and the median number was 678 . All cells were washed and fixed in a final concentration of 2% paraformaldehyde prior to analysis . Cell sorting was performed on a MoFlo Astrios from Beckman Coulter . All flow experiments performed at Boston University School of Medicine Flow Cytometry Core Facility . Cell activation and phenotypes were determined by CD69 expression ( Brilliant Violet 421 anti-human CD69 antibody; Clone FN50 , BioLegend ) and CCR7 and CD45RA expression ( Pe/Cy7 anti-human CCR7 antibody; Clone G043H7 , BioLegend and PerCP/Cy5 . 5 anti-human CD45RA antibody; Clone HI100 , BioLegend ) . We had a minimum of 1000 events to be included as a data point in Fig 2 . Primary CD4+ T cells were transduced with CARs and allowed to return to a resting state for 1 week prior to cell sort . Cells were then stimulated overnight with plate-bound Her2 as described above . Untransduced CD4+ T cells were left unstimulated or were plated in a solution of 1 μg/mL CD28 ( Mouse Anti-Human CD28 , #555725 , BD Biosciences ) on previously coated wells of 1 μg/mL CD3 ( Mouse Anti-Human CD3 , #555329 , BD Biosciences ) . We used a minimum input of 2 . 5x104 primary cells per experimental condition . Cells were then washed in PBS and lysed for RNA extraction using Qiagen miRNeasy Mini Kit ( #217004 ) . Microarrays and statistical support were provided by BU Microarray and Sequence Resource Core Facility . cDNA was made and samples were run on a Human Clariom S Array . Human Clariom S CEL files were normalized to unstimulated cells to produce gene-level expression values using the implementation of the Robust Multiarray Average ( RMA ) [54] in the affy package ( version 1 . 36 . 1 ) [55] included in the Bioconductor software suite ( version 2 . 11 ) [56] and an Entrez Gene–specific probe set mapping ( 21 . 0 . 0 ) from the Molecular and Behavioral Neuroscience Institute ( Brainarray ) at the University of Michigan [57 , 58] . Array quality was assessed by computing Relative Log Expression ( RLE ) and Normalized Unscaled Error ( NUSE ) using the affyPLM package ( version 1 . 34 . 0 ) . Analyses of variance were performed using the f . pvalue function in the sva package ( version 3 . 4 . 0 ) . Differential expression was assessed by performing Student's t test on the coefficients of linear models created using the lmFit function in the limma package ( version 3 . 14 . 4 ) . In this way , a one-way ANOVA p value was obtained using a linear mixed effects modeling approach to account for differences between donors . Correction for multiple hypothesis testing was accomplished using the Benjamini-Hochberg false discovery rate ( FDR ) [59] . All microarray analyses were performed using the R environment for statistical computing ( version 2 . 15 . 1 ) . All genes with FDR q values below 0 . 01 were plotted on a heatmap and arbitrarily separated into 5 clusters based on expression profiles . Determination of gender was based upon log2 expression of the Y-linked genes DDX3Y , KDM5D , RPS4Y1 , USP9Y , and UTY . 4x105 Jurkat cells were washed and lysed for luciferase analysis 24 h post infection , while 4x105 primary T cells were measured at 4 days post infection . Luciferin ( Promega ) was added and luciferase activity was measured via BioTek Synergy HT Microplate Reader . A minimum input of 8x105 cells per experimental condition was lysed in Tris-EDTA buffer prior to Alu-PCR . Nested PCR strategy was adapted from Agosto et al . , 2007 [20] . Briefly , integrated HIV-1 DNA was amplified using forward primers for the luciferase sequence and reverse primers for human Alu ( see S1 File ) . The first reaction was performed on a TProfessional Thermocycler from Biometra according to the following conditions: 4 m at 95° followed by 20 cycles of 15 s at 93°C , 15 s at 50°C , and 2 . 5 m at 70°C . A second round of amplification was then performed using a forward primer , a reverse primer , and a probe for real time PCR within the HIV-1 3’ R / U5 region ( see S1 File ) . The amount of amplified copies of HIV-1 was determined based on an NL4-3 plasmid copy standard . The second reaction was performed on an Applied Biosystems QuantStudio 3 Real-Time PCR system with heating for 4 m at 95° and real-time PCR conditions of denaturation for 15 s at 95°C , annealing for 30 s at 60°C , and extension for 1 m at 72°C . We lysed 1 . 5x105 cells per experimental condition in TRIzol Reagent ( Invitrogen ) . RT-PCR for HIV-1 mRNA was performed using forward primers and reverse primers for unspliced HIV-1 tat , and all values were normalized against beta-actin as a housekeeping gene ( see S1 File ) . The reaction was performed on an Applied Biosystems QuantStudio 3 Real-Time PCR system with heating for 15 m at 94°C and real-time PCR conditions of denaturation for 15 s at 94°C , annealing for 30 s at 60°C , and extension for 30 s at 72°C . 5x106 Jurkat cells were infected with NL4-3 . Luc . ChIP was performed 24 h later according to Natarajan et al . , 2013 [47] with the addition of a nuclei isolation step using Farnham Lysis Buffer prior to sonication with a Bioruptor Pico . Specific details are listed in S1 File . Antibodies used included anti-NELF-d ( Antibody TH1L from Proteintech Group ) , anti-RNA Polymerase II antibody ( Clone N20 from Santa Cruz Biotechnology ) , anti-histone H3 antibody ( Product 06–599 from Millipore Sigma ) , and Normal Rabbit IgG ( Product 12–370 from Millipore Sigma ) . Primers used for the transcriptional start site include the forward primer at +30 and the reverse primer at +239 . Primers used for transcriptional elongation include the forward and reverse primers within the tat gene ( see S1 File ) . Except for microarray analysis detailed above , all statistical analysis performed using unpaired Student’s t test with significance thresholds of *p<0 . 01 , **p<0 . 001 , and ***p<0 . 0001 . Because our experiments were performed on a sample population under the same conditions , we assumed that our data would be normally distributed . In agreement with this assumption , all experimental data points were less than 2 standard deviations from the mean . Where appropriate , normality was tested with a Shapiro-Wilk test . | Activation of CD4+ T cells facilitates HIV-1 infection; however , whether there are minimal signals required for the establishment of infection , replication , and latency has not been explored . To determine how T cell signaling influences HIV-1 infection and the generation of latently infected cells , we used chimeric antigen receptors to create a tunable model . Stronger signals result in robust HIV-1 expression and an inducible latent population . Minimal signals predispose cells towards latent infections that are refractory to reversal . We discovered that repression of HIV-1 transcription immediately after infection is due to RNA polymerase II pausing and inefficient transcription elongation . These studies demonstrate that signaling events influence the course of HIV-1 infection and have implications for cure strategies . They also provide a mechanistic explanation for why a significant portion of the HIV-1 latent reservoir is not responsive to latency reversing agents which function by modifying chromatin . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
]
| [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"engineering",
"and",
"technology",
"pathogens",
"signal",
"processing",
"immunology",
"microbiology",
"dna",
"transcription",
"retroviruses",
"viruses",
"immunodeficiency",
"viruses",
"immune",
"receptor",
"signaling",
"rna",
"viruses",
"membrane",
"receptor",
"signaling",
"tcr",
"signaling",
"cascade",
"immune",
"system",
"proteins",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"medical",
"microbiology",
"hiv",
"gene",
"expression",
"microbial",
"pathogens",
"t",
"cells",
"hiv-1",
"biochemistry",
"signal",
"transduction",
"t",
"cell",
"receptors",
"cell",
"biology",
"viral",
"pathogens",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"immune",
"receptors",
"lentivirus",
"cell",
"signaling",
"organisms",
"signaling",
"cascades"
]
| 2019 | Strength of T cell signaling regulates HIV-1 replication and establishment of latency |
Nutrient acquisition is a critical determinant for the competitive advantage for auto- and osmohetero- trophs alike . Nutrient limited growth is commonly described on a whole cell basis through reference to a maximum growth rate ( Gmax ) and a half-saturation constant ( KG ) . This empirical application of a Michaelis-Menten like description ignores the multiple underlying feedbacks between physiology contributing to growth , cell size , elemental stoichiometry and cell motion . Here we explore these relationships with reference to the kinetics of the nutrient transporter protein , the transporter rate density at the cell surface ( TRD; potential transport rate per unit plasma-membrane area ) , and diffusion gradients . While the half saturation value for the limiting nutrient increases rapidly with cell size , significant mitigation is afforded by cell motion ( swimming or sedimentation ) , and by decreasing the cellular carbon density . There is thus potential for high vacuolation and high sedimentation rates in diatoms to significantly decrease KG and increase species competitive advantage . Our results also suggest that Gmax for larger non-diatom protists may be constrained by rates of nutrient transport . For a given carbon density , cell size and TRD , the value of Gmax/KG remains constant . This implies that species or strains with a lower Gmax might coincidentally have a competitive advantage under nutrient limited conditions as they also express lower values of KG . The ability of cells to modulate the TRD according to their nutritional status , and hence change the instantaneous maximum transport rate , has a very marked effect upon transport and growth kinetics . Analyses and dynamic models that do not consider such modulation will inevitably fail to properly reflect competitive advantage in nutrient acquisition . This has important implications for the accurate representation and predictive capabilities of model applications , in particular in a changing environment .
Nutrient transport ( e . g . , of NO3- , NH4+ , PO43- ) typically occurs via secondary active porters that are either matched for a specific nutrient molecule type , or for similar types [12]; thus a transporter for NO3- will not transport NH4+ , while similar amino acids such as the cationic group arginine , lysine , histidine and ornithine may share the same transporter [13] . In addition , individual nutrient types may be taken up by several different transporter proteins [14–16] , some of which may support biphasic kinetics [16–18] . Here , to simplify discussions , we will consider transport via a single ( monophasic ) transporter type . While transporter proteins are not strictly enzymes ( as they typically do not change the chemical form of their substrate ) , they express an affinity for the nutrients they transport; by analogy with the Michaelis-Menten half saturation value of enzymes , KM , we term this substrate concentration KT . The constant KM is a function of the affinity of the enzyme for the substrate in classic Michaelis-Menten terminology and is determined assuming that all factors other than substrate availability are non-limiting . Determining KT is more complex because transporter functionality depends on the integrity of the membrane in which the transporter proteins function , ionic gradients generated by primary active transporters required to support the operation of the typically secondary-active nutrient-transporters , as well as on the aforementioned absence or presence of short and longer term feedback processes modulating transport itself into the functional cell . Another defining criterion for enzyme functionality is the maximum level of activity , kcat , which is described in units of mole of substrate consumed ( or product given ) per mole of enzyme per unit of time ( Table 1 ) . The maximum rate of enzyme activity in a given sample of biological material , which is a product of kcat and the concentration of enzyme protein , sets the value of the maximum process rate , Vmax , in Michaelis-Menten kinetics . It is important to note that the amount of enzyme in an assay does not affect the value of KM , while the value of Vmax in the assay is linearly related to enzyme concentration . The value of Vmax can thus be seen as being somewhat ambiguous , only being useful for a specific assay incubation . For considerations of whole-organism physiology , the value of kcat needs to be placed in the context of the total demand for its activity , the size ( mass ) of the enzyme and thence for the total resource expenditure for that enzyme within a given cell ( e . g . , for such calculations applied to the enzyme fixing CO2 , RuBisCO [19] ) . The maximum rate of activity in a given cellular system ( Tmax ) is analogous to Vmax in an enzyme assay . Accordingly , while the value of KT is independent of the number of transporter proteins in the cell , the value of Tmax is indeed dependent on that number . The extent to which Tmax exceeds Gmax , noting that transporter activity is modulated by post-transport physiology , helps to explain why KG is lower than KT , as illustrated in S1 Fig and the adjoining online text . In reality there are many hundreds if not thousands of transporter proteins in operation across the plasma-membrane of an individual cell . Theoretical estimates of relative nitrate and phosphate transporter density suggest that a specific transporter type will generally cover less than 0 . 1% of the cell surface under nutrient limited conditions [20] . The number of transporter proteins , and hence the maximum rate of nutrient transport ( Tmax ) , also varies greatly with the nutritional status of the cell and for different nutrients , with ammonium transport and assimilation being much faster than for nitrate [21] . An example of the differences between ammonium and nitrate transport potential , and concurrent needs of N assimilation at different levels of N-stress is given in S2 Fig . The linkage between nutrient transport and assimilation , and ultimately growth , is modulated via the expression of transport capacity for specific nutrient types via end-product ( de ) repression signals . These events involve both short-term control , for example satiation feedback regulation upon the operation of existing transporter proteins , and longer-term control via synthesis and removal of transporter proteins . This feedback occurs more quickly following ammonium and than nitrate transport because of the rapidity of both ammonium transport and of its assimilation [6 , 22] . Nitrate may also be accumulated in larger cells further decoupling processes of N-assimilation from transport . Thus , depending on the organism size and nutrient status , the nutrient being tested , experiment sampling , and subsequent data processing methodology , the values of both Gmax and KG may differ significantly from Tmax and KT [6] . Experiments using a given species and nutrient , for example varying the period of N-limitation , may likely give useful information on trends . However , interpretations of inter-species and inter-nutrient differences in Umax and KU , especially when derived by different researchers , carry a high degree of uncertainty . Estimates of KT for nutrient transport are very rare , and values for phytoplankton nutrient transporters are rarer still [23] , but a value in the range of 0 . 5–2 μM has been reported [15] . In the following we will assume KT = 1μM . For comparison , the KM for enzymes processing biochemical transformations are typically in the mM range [24] . Just as the importance of the numeric value of Vmax needs to be placed in the context of the enzyme sample in which it has been measured , so the value of Tmax needs to be placed in the context of the cell in which it is located . The value of Tmax may be expressed per cell , as a specific transport rate either following N-source uptake using 15N , or as a C-specific rate ( this is shown in S2 Fig ) . Nutrient availability for the cell does not just reflect the bulk water nutrient concentration ( S∞ ) , which is readily measured , but it reflects the interactions between processes adding and removing nutrient molecules around the individual cell which affects the substrate concentration ( S0 ) at the transporter protein . Thus S0 is also affected by turbulence , cell size and the cell’s motion [25–27]; collectively these determine the formation of a boundary layer around the cell and thence affect diffusion to the sites of transport . Cell size is a critical determinant in transport kinetics , as it affects the boundary layer thickness and hence the relationship between S0 and S∞ . It is thus constructive to express Tmax in the context of the surface area of the plasma-membrane in which the transporter proteins reside . If we assume a spherical cell form , with a given equivalent spherical diameter ( ESD ) and an equal distribution of transporter proteins over the membrane surface , we can then report Tmax in terms of a transport rate density ( TRD; Table 1 ) . Thus , for the transport of ammonium-N , units of TRD would be g ammonium-N d-1 μm-2; that is to say that every day across every μm2 of cell plasma-membrane area so many g ammonium-N could be transported assuming no satiation feedback . The value of TRD is enabled by the activity of many transporter proteins spread over the cell surface area ( SA ) , each of which has its own KT and kcat . TRD is thus TRD=TmaxSA ( 2 ) and Tmax is Tmax=kcat∙{transporterproteinspercell} ( 3 ) The larger the cell , the greater the surface area but there is no reason to necessarily expect the value of TRD to differ according to cell size . In the following we ignore changes in cell size associated with nutrient availability ( e . g . N-limited cells are typically smaller , while P-limited cells are larger ) and environmental conditions ( e . g . growth at different temperatures and irradiance [28] affect cell shape and size ) . Growth itself is not a simple function of the presence of external nutrient availability ( even if estimated more accurately as S0 rather than S∞ ) , but is primarily a function of availability of that nutrient within the cell , and the allied biochemical processes associated with its assimilation into biomass . We thus need to consider transport rates in the context of supply and demand for the cell . Depending on the nutrient status , the value of Tmax changes ( S2 Fig ) , and consequentially so does the value of TRD . We can now define two important specific values of TRD . These are the values of TRD needed to enable G = Gmax ( TRDGmax ) , and the maximum possible value of TRD ( TRDmax ) ; the latter defines the value of TRD which aligns with the absolute maximum value of Tmax ( Tabsmax ) which is usually expressed by a cell at an intermediate level of nutrient stress ( S2 Fig ) . By analogy with the plots shown in S2 Fig , we can also consider the excess in transport potential that develops during nutrient stress as the ratio of TRDmax: TRDGmax ( δTRD; Table 1 ) . At saturating concentrations of nutrient and plausible maximum growth rates we can assume that diffusion is not limiting the supply of substrate to the transporter proteins ( S0 ≈S∞ ) , and hence we can estimate the value of Tmax ( as per S2 Fig ) and hence TRD . From experimental work for ammonium and nitrate transport into the coccolithophorid Emiliania huxleyi , raphidophyte Heterosigma carterae and the diatom Thalassiosira weissflogii we compiled the data shown in Table 2 . These values exploit relationships between cell biovolume measured using an Elzone ( Coulter counter–like ) instrument , and C-biomass derived from elemental analysis . In Table 3 , comparative values for TRD are presented , calculated using the allometric relationships of cell size to C-content taken from the literature [9] . While there are significant differences between the C- , and thus the N-content of the cells computed according to these different methods , from these estimates we obtain a feel for a likely maximum value of TRDmax . For a given computational choice ( Table 2 or Table 3 ) the value of TRDmax is not so different between organisms of markedly different taxonomy , size and maximum growth rate potential . These values suggest a decreasing scope for excess in transport potential δTRD ( i . e . , TRDmax: TRDGmax ) with increasing size , which may be expected , given the associated changes in surface area to volume ( SA:Vol ) ratio . An analysis of the data compiled by [11] , which reports experimentally derived nutrient uptake maxima , and assuming Umax = Tmax , yields average TDRmax that are broadly in line with those in Tables 2 and 3 . Those data yield TRD values ( as pg nutrient μm-2 d-1 ) of 0 . 075 ( +/- SE 0 . 041 ) , 0 . 115 ( +/- SE 0 . 0137 ) and 0 . 172 ( +/- SE 0 . 069 ) for ammonium-N , nitrate-N and phosphate-P uptakes , respectively , with no statistical relationship with ESD . It is noteworthy that the TRD values for ammonium estimated from the data compiled by [11] are half those for nitrate; ammonium Tmax and thus TRD is expected to be much greater than the values for nitrate transport [21 , 34] , which could indicate confounding estimation of kinetic parameters by different researchers , as explained earlier . Ultimately the balance of supply and demand is reflected in how close an organism comes to attaining its maximum growth rate , Gmax . It is this maximum rate of growth , and the form of the functional curve relating nutrient concentration to the achieved growth rate ( G ) that help define competitive advantage , and certainly do so in simple mathematical models . However , while the performance of each transporter protein may be expected to conform to the RHt2 equation of Michaelis-Menten kinetics , diffusion limitation is expected to decrease potential transport at lower nutrient concentrations [35 , 4 , 36] , and the satiation feedback is expected to suppress transport rates at higher concentration ( S2 Fig ) . In short , there are various reasons to expect that a RHt2 response curve ( as used in simple models ) will not well describe the true functional response curve between the bulk nutrient concentration ( S∞ ) and G . Indeed , we should likely not expect such a RHt2 relationship even between S0 and G ( S1 Fig ) . Let us now consider the situation that aligns with a growth rate at half the maximum value ( G0 . 5 ) . At this rate , the residual steady-state nutrient concentration in the bulk medium ( S∞ ) would equate to the half saturation value for growth , which defines KG . The value of Tmax in cells growing at the N-status equal to G0 . 5 is much higher than the value of Tmax in cells growing at Gmax ( S2 Fig; e . g . [21] ) . In addition , the amount of N required to support growth at G0 . 5 is less than that required to support G = Gmax . If , for example , we consider Gmax to be associated with a maximum cellular N:C ( g:g ) of 0 . 2 , and G = 0 with a minimum N:C of 0 . 05 , then a cellular N:C aligning with G0 . 5 would be expected to be ca . 0 . 125 gN gC-1 ( S2 Fig ) . In such a situation , the potential excess ( δTRD ) in transport capacity , of Tmax , at G0 . 5 could be ca . 20 fold the nutrient transport rate required at Gmax . It is thus readily apparent that cells with different stoichiometries will exhibit different growth kinetics with respect to nutrient concentration , all else being the same . There is one other important part of the jigsaw , and that concerns the relationships between cell size , the cellular carbon density as affected by vacuolation , and cell shape . For simplicity we assume a spherical cell , which then sets surface area ( SA ) as a simple geometric function of cell size ( ESD ) . Vacuolation in protists , and especially in diatoms , increases markedly with ESD [9 , 37] , and hence the demands for nutrient transport across each μm2 of cell surface does not simply relate to cell size . Having described the physiological framework , and considered the experimental data , we now proceed to extend the analysis according to allometric constraints across a range of sizes , organism types and motilities . The questions that we consider are: The emphasis here is on factors that impact upon KG , namely S∞ that support G0 . 5 . This value of KG can be seen to be an emergent property of TRD , KT , cell size , Gmax , cell motility , cell vacuolation and cellular elemental stoichiometry . To our knowledge , no previous study has considered the interconnected nature of all these facets . Collectively these also embrace the core features considered in classic trait trade-off studies .
Fig 1 shows the potential growth rate at given external bulk nutrient concentrations ( S∞ ) in terms of dissolved inorganic-N , DIN , for different cell types and configurations , all with the same fixed maximum growth rate of Gmax = 0 . 693 d-1 . These plots clearly show the competitive advantage for nutrient transport of being small , and of motion achieved through either swimming ( flagellated phototrophic protist ) or sedimentation ( diatom ) . Thus the value of S∞ supporting G0 . 5 ( G = 0 . 693/2 = 0 . 346 d-1 ) , which is the value of KG , decreases with cell size and with motion . At cell ESD below 5μm , at this growth rate , nutrient concentrations at the cell surface are similar to those in the bulk water . The cellular carbon density also has an important impact on the growth-nutrient kinetics; increasing vacuolation with size ( for a given C:N stoichiometric configuration ) decreases the requirement for N transport . It is thus apparent that diatoms can compensate significantly for increasing cell size through being more vacuolate and hence having de facto a lower than expected SA: cell-N ratio compared to a typical protist phytoplankton . While altering the value of KT ( assumed by default as 1μM ) changes KG , the relationship is not pro rata; thus halving KT decreases KG to ca . 75% , and doubling KT increases KG to ca . 150% . In Fig 2 , values of KG obtained with different cell configurations growing with different maximum growth rates are plotted , showing that smaller cells can attain a higher Gmax relative to KG; their value of Gmax/KG is higher . Fig 3 also shows the potential for cell motion and/or cellular carbon density to compensate for the negative impact of increasing ESD . For a given cell configuration , however , the value of Gmax/KG is invariant with changing Gmax ( Fig 3 ) . The negative relationship between Gmax/KG and ESD varies strongly between cell configurations , and becomes more variant between configurations at larger ESD ( Fig 4 ) . The power slopes between Gmax/KG and ESD are given in Table 4; assuming a cellular carbon density that is fixed ( C150 ) , or accords with a generic protist phytoplankton ( Cprot ) or with a diatom ( Cdiat ) . More details regarding the organism’s configuration are given in Table 1 . The slope exceeds -1 . 5 , but motility ( through swimming or sedimentation ) and increasing vacuolation with ESD mitigate the slope to closer or less than -1 . To consider the implications of variable elemental stoichiometry , Fig 5 presents the relationship between cell size and the minimum N:C quota ( NCmin ) and the nutrient concentration that half saturates transport of dissolved inorganic-N for protists ( non-diatoms ) that are motile or non-motile . This assumes a fixed maximum growth rate and fixed maximum N:C quota . These plots demonstrate a linear increase in KG as the difference between NCmax and NCmin decreases; cells with a more restricted N:C quota need more N and thence are disadvantaged if DIN acquisition is the sole limiting factor . S3 and S4 Figs show how N-specific transport ( which aligns with growth rate ) varies with nutrient concentration for cell configurations Cprot and Cdiat , considering different maximum growth rate potentials , ESD , and different relationships between N-status and Tmax . These plots show how the difference between bulk water and cell surface nutrient concentrations ( S∞ vs S0 ) for a given transport rate increases with ESD and with maximum growth rate . Also apparent is that , for a given KT ( all these plots assuming the same value of 1 μM ) the relationship between N-status and Tmax has a very significant effect on the kinetics ( as expected from S1 Fig ) . To consider whether these kinetics could be adequately described through application of a simple RHt2 response curve ( as per Eq 1 ) , such a curve form was fitted to the model output using an iterative approach ( as supported by SigmaPlot 12 . 5 ) ; the fit assumed either a free maximum rate , or a maximum rate that is fixed equal to the value of Gmax . Especially notable , where Tmax increases with deteriorating N-status ( Fig 6 ) , is that the form of the response curve appears steeper and/or plateaus more abruptly than for a RHt2 curve ( S3–S6 Figs ) . Nonetheless , the R2 values for all of these fits exceed 0 . 98 . The RHt2 plots typically overestimate transport at nutrient concentrations aligning with the value of KG and could significantly over-estimate ( free-fitting maximum; “RHt2” plots in S3–S6 Figs ) or under-estimate ( plateau fixed equal to Gmax; “RHt2 fGmax” plots in S3–S6 Figs ) transport at higher nutrient abundance . Rather than using simple hypothetical relationships between N-status and Tmax ( S3 Fig and S6 Fig ) , in Fig 7 experimentally derived response curves ( from S2 Fig ) were deployed . Again , the importance of the form of the relationship between N-status and Tmax is clear; especially for the nitrate curves , the deterioration in transport capacity at low N-status ( low N:C in S2 Fig ) results in the cell-surface nutrient values being closer to the bulk water values than may otherwise have been expected . Fig 7 also shows how RHt2 curves that give statistically acceptable fits also give differences in projected transport rates for a given nutrient concentration that could be significant in simulations . This is especially so for nitrate-supported growth .
We may consider that transporter proteins are specialist enzymes . There is an established literature exploring the competitive advantages , and evolution , of enzymes of different kcat and KM . Pettersson ( 1989 ) [42] considers the evolution of the value of kcat/KM noting that , beyond the initial phase that sees the expected increase in kcat and decrease in KM , enzyme evolution displays a linked increase in both kcat and KM; the value of kcat/KM approximates the diffusion control limit at the level of the enzyme molecule . Several studies [43–45] discuss the usefulness of this so-called “specificity constant” ( kcat/KM ) pointing out various problems both with the usefulness of the value itself , and with its some-time alternative title as a value of “catalytic efficiency” . Interpretations of transporter kinetic parameters , operating at the site of individual transporter proteins , would be similarly implicated in such considerations . Just as trying to piece together whole organism biochemical evolution through reference to kcat/KM for all the constitutive enzymes in an organism is fraught with problems [42] , so too are considerations of transport kinetics for different substrates into different species . However , it is noteworthy that our analysis indicates that , for a given cell configuration ( size , motility , value of Ccell , stoichiometry; Fig 6 and Fig 7 ) , the value of Gmax/KG is constant , as is kcat/KM expected to be constant in an evolutionary mature enzyme . The phytoplankton literature has hitherto explored the relative competitive value of organisms under nutrient limitation through reference to ( in our terminology–see Table 1 ) to Umax/KU . This value of Umax/KU has been termed “affinity” in parts of this literature [10 , 46] . Such usage of “affinity” conflicts with traditional parlance for enzyme affinity , which defines affinity by just the half saturation constant KM . The form and interpretation of Umax/KU is also different to that for kcat/KM for enzymes; while KU may approximate to KM , Umax is de facto a function of the product of transporter kcat and the number of transporter proteins . The number of transporter proteins varies with cell size , nutrient status and likely also with Gmax . In addition , there is the practical challenge of measuring Umax , being as it is a function of Tmax ( the rate of transport at the start of the experimental incubation , at t0; [6] ) and incubation conditions during the assay . In consequence , the values of Umax and KU , and thence of their ratio , are subject to various confounding issues . The value of Umax/KU could , under ideal conditions of measurement , perhaps be equated to Tmax/KT; however , there is still the question as to the impact of nutrient status upon Tmax ( S2 Fig ) , and the complication that KT is the substrate value at the transporter protein ( S0 ) while KU is the value of the substrate concentration in the bulk medium ( S∞ ) . The underlying explanations and potential trade-offs in expression of the uptake affinity defined as Umax/KU has been argued to lack a mechanistic basis , hence leading to a potential misrepresentation of primary production in modelling approaches [3 , 47 , 5] . Our results indicate why a search for such a mechanistic basis has proven so difficult; there are too many confounding factors . An alternative approach considers nutrient uptake as a function of cell traits and actual nutrient availability in a turbulent environment [4 , 48 , 49] . The non-linear formulation describes so-called affinity as a function of cell size , density of uptakes sites at the cell surface ( i . e . transporter proteins ) and turbulence [5] . This diffusion- limited nutrient uptake results in a linear scaling of affinity with the cell diameter or radius ( r ) . While some experimental results are consistent with this scaling [50] , the general picture drawn by laboratory experiments over a wide range of sizes of taxa indicate a scaling closer to the square of cell radius [10 , 51] that is with the cells surface area , a trend that becomes more pronounced with decreasing cell size . Theoretical arguments have suggested that this mismatch might stem from the fact that cells are not “perfect sinks” , hence are not able to absorb all nutrients at the cells surface immediately as assumed by diffusion limited nutrient uptake [20] , which is likely once satiation feedback develops . According to these considerations , while smaller cells are favoured by a larger surface to volume ratio , they also require a higher transporter density to achieve maximum affinity and would thus have higher relative investment costs [20] . However , Tmax increases during at least the initial phase of nutrient-limitation ( S2 Fig ) , which demonstrates an increased synthesis cost for transporters in such nutrient-limited cells; this suggests that the investment cost in transporters is not significant . There are clearly challenges with all the above analyses , centring upon what exactly Umax and KU index as curve-fitting parameters for RHt2 curves fitted through imperfect ( and only partially understood ) experimentally-derived data . With suitable methods , estimates of Umax will approach Tmax , and estimates of KU will approach KT [6] . The numeric disparity between these variables depends on the nutrient status of the cell , the size of the cell ( and thus how close S0 is to S∞ ) , the form in which the nutrient is available , and the capacity of the cell to accumulate unaltered that particular nutrient prior to the development of satiation feedback . In consequence , greater challenges could be expected when measuring the kinetics of ammonium transport , which is assimilated very rapidly [8] and not accumulated . The ability of the diatom Phaeodatylum to take up the un-metabolisable ammonium analogue methlyamine is many orders of magnitude higher than for any other N-source [21] . This likely reflects the fact that methylamine entering via the ammonium transporter is not subject to the usual very rapid accumulation of the ammonium-transport-repressor signalling amino acid glutamine [8] . Lesser problems can be expected when measuring nitrate transport into a large vacuolated diatom that may accumulate nitrate [52] , in comparison to transport into a nanoflagellate that lacks such vacuoles . It may therefore likely be no coincidence that the ( few ) data for kinetics for ammonium transport collated by Edwards et al . ( 2015 ) [11] appear so competitively poor in comparison with those for nitrate when the converse might have been expected . Similarly we expect fewer challenges when measuring phosphate transport into a cell type that accumulates polyphosphate . Nutrient “affinity” [10 , 46] , which has been described in our terminology as Umax/KU , has typically not been related to the C:N:P stoichiometry of the cell nor to the cellular carbon density both of which will affect the numeric value of this index . Together , these additional data would provide links between nutrient-status and Tmax and to the level of vacuolation affecting resource demand to be satisfied by transport over the cell surface . Collectively , stoichiometry ( Fig 5 ) and cellular carbon density ( Fig 1 ) affect the cell’s demand for the nutrient , which is a critical factor affecting the relative importance of any index of nutrient affinity . There is , however , scope for Tmax to vary allometrically on account of the packing of transporter proteins within the plasma membrane ( Fig 6 ) ; which is consistent with the suggested explanation of the discrepancy between theoretical scaling and observed values of Umax/KU [20] . Further , and of greater significance for large non-diatoms protists than for diatoms , there is scope for the maximum growth rate to be limited by TRD attaining TRDmax ( Fig 8 ) . That is , if TRDmax = TRDGmax there is no scope to further enhance transport during nutrient stress . This is important because the value of KG is a function of the potential transport over the required capacity in transport ( S1 Fig ) , as the ratio Tmax: TGmax . This means that larger cells , and faster growing cells of a given configuration ( cell type and motility ) , are expected to have a higher KG . There are also additional features of ecophysiology that affect the medium term dynamics of nutrient transport . There is for example a difference in the handling of ammonium versus nitrate , that sees the uptake and assimilation of ammonium more constrained to just the light phase . Thus ammonium transport rates during light may have to be double those expected looking at the day-average value , while nitrate assimilation is more likely split over the whole day [30] . In these contexts , it is interesting to note the relationships between ESD and Gmax for different cell types [53] , and that the typical value of Gmax in phytoplankton equates to a division per day ( Gmax = 0 . 693d-1 ) , aligning with RuBisCo activity [19] . It is not just nutrient acquisition at nutrient-limiting concentrations that may be limiting growth rate potential; maximum transport at non-limiting concentrations may also be a factor ( Fig 8 ) . While for nitrate transport , there may be the potential for the expression of high-rate transporters , endowing the cell with a biphasic kinetic capability [18 , 54 , 55] , this may be less likely for ammonium transport . Ammonium is highly toxic at high internal concentrations and its transport appears , unsurprisingly , tightly regulated . Ammonium is also normally present at low ( often at vanishingly low ) concentrations in natural waters , as the product of N-regeneration in ecosystems with low inorganic N concentrations . If for a given cell , the ammonium transporter exists only as a high affinity system , which is incapable of supporting growth at the highest rates because of limitations in TRD for ammonium , then high growth rates in large protists may only be possible when augmented by nitrate transport . This would place an interesting new spin on our understanding of ammonium-nitrate interactions , with implications for modelling biogeochemical and ecological events . The results of our analysis show how features relating to the regulation of the synthesis and kinetics of transporter proteins , as well as to stoichiometric and allometric features of the cell , all play a part in the story . Arguably , the competitive advantage of an organism would be best indexed by the value of Gmax/KG as this integrates over all aspects of the organism’s nutrient physiology . We thus emphasise factors affecting KG . In the following we assume for the most part that all else remains constant ( i . e . KT , TRDmax , NCmax and NCmin are constant ) and consider the impacts of each of these factors upon the system . If the cellular carbon density is constant across cell sizes , then there is a clear and powerful impact of cell size on KG ( Figs 4–7 ) . Smaller cells are much better equipped than are larger cells in this regard; this is because the SA:Vol ratio directly translates to a SA:N-demand ratio , as well as to lower diffusion limitations in smaller cells [56] . However , in reality there is an important allometric relationship between cellular carbon density and cell size [9] such that larger cells have a lower cellular carbon density . For diatoms in particular , which are increasingly vacuolate at large size [37] , this greatly decreases the needs for nutrient transport across a given area of plasma-membrane . According to the calculations presented here , large diatoms with high sedimentation rates appear potentially to be much better adapted to make use of low nutrient concentrations than one may expect if one was to assume a fixed cellular carbon density ( i . e . , Cdiat vs C150 ) ( Fig 4 ) . The consequences of this decrease in cellular carbon density with cell size is actually secondary to the decrease in N-cell density; the above mentioned mitigation of cell size on KG in consequence of the lower N-cell density thus assumes that cell stoichiometry is the same . From the effects of altering the range of cell stoichiometry , shown in Fig 5 , we conclude that cell stoichiometry and the form of the relationship between stoichiometry and growth rate ( the quota relationship–see [57] ) are also important factors to consider when reviewing calculations of KG . That is to say , if larger cells had a high NCmin , such that the value of N:C at G0 . 5 was elevated , then the mitigation afforded through being more vacuolated would be eroded . Conversely , if smaller cells were relatively N-rich , then the advantage of being small would be eroded . For example cyanobacteria are typically relatively N-rich [58] and would therefore not be so competitive as may at first appear . The physiology of nutrient acquisition and stoichiometry has the potential to override , or at least partially compensate for , limitations at transport [59] . Models considering detailed explorations of nutrient uptake kinetics thus need also to relate those kinetics to variable stoichiometry and cell size , and not assume simple fixed relationships . For phosphate transport , as for ammonium transport , TRDmax is likely very much higher than TRDGmax . In addition , the strongly curved form of the P:C quota relationship [57] will also have a strong impact upon KG for P-limited growth as the P:C value in cells growing at G0 . 5 will be low . Our analysis suggests that for smaller cells ( ca . <5μm ESD ) motion has little additional scope to moderate diffusion limitation . Above that size , the negative effect of size is greatly countered ( though not negated ) by motion through swimming or sedimentation ( Figs 1–5 ) . Note that sedimentation is affected directly by Stokes law; hence differences in cell mass between species , and with nutrient status may affect sedimentation rates [60] . While it may be tempting to explain motility primarily as a mechanism to enhance competitive advantage for nutrient transport ( i . e . , through lowering KG ) , the role of motility is also related to behaviour linked to vertical migration [61 , 62] . Motility is also important for finding prey to support mixotrophy , an activity present in even the very smallest flagellated species , with an ESD of <3 μm , Micromonas [63] . Sedimentation in diatoms is a common trait [64 , 65] , often considered as detrimental but having clear advantages for nutrient acquisition at low concentrations in turbulent water systems ( Fig 4 ) . For diatoms , sedimentation adds significantly to the advantage of becoming increasingly vacuolate with larger ESD ( Figs 4–7 ) . Given that cell size usually also confers an anti-predator advantage , this means that larger diatoms appear better adapted to dominate in turbulent waters ( in which their sedimentation de facto confers motility ) than may otherwise appear . Our analysis indicates that the relationship between Gmax/KG and Gmax is flat for a given ESD ( Fig 5 ) . This relationship is useful as it permits the estimation of KG for a given organism type , motility and size . It also means that a given organism will have a lower KG if its Gmax should decrease through adaptation , or indeed through acclimation , to different environmental conditions . The analysis also indicates that there is scope for a much greater spread in nutrient-related kinetics in larger cells ( Fig 4 ) . For smaller cells there is less effect of motility , and less variation in cell-C density; inter-species variation will thus generate increasing “noise” in the relationship between ESD and kinetics in larger cells . The value of Gmax/KG reflects many interactions and as a summary parameter provides an index for competitive advantage in simple nutrient-competition ( bottom-up controlled ) systems . The value of KG itself also has important implications for the health of the cell; it defines the bulk water nutrient concentration ( S∞ ) supporting a state of health aligning with G0 . 5 . Health affects the intrinsic mortality rate of the cell , a factor that is typically not included in models scaled to nutrient status , but one that is important as a selective feature [66 , 67] . A poor health status adversely affects the operation of repair mechanisms , e . g . compensating for photo-damage [68] , and explains the duration of the lag phase of growth seen when nutrient-starved microalgae are re-fed [69] . Simple models relate nutrient concentration to transport rate and thence to growth rate using a rectangular hyperbolic type 2 ( RHt2 ) response curve , in line with Monod ( 1949 ) [70] . From our analysis ( Figs 9 , S1 and S3–S6 ) RHt2 cannot be expected to well define the actual relationship between nutrient concentration in the bulk medium ( S∞ ) and transport . The fitting of RHT2 tends to over-estimate transport at lower nutrient availability and over or under estimate it at high availability . The expected relationship plateaus more abruptly than RHt2 can describe it . It is noteworthy that the fit of RHt2 to the modelled relationships was high ( R2 > 0 . 95 in all instances , and most > 0 . 98 ) ; the “noise” in biological measurements that is inherent in experimental procedures of transport and growth rates [6] will inevitably result in a statistically acceptable fit to RHt2 . Nonetheless , RHt2 does not appear to be appropriate , and the apparent subtle differences in the form of the described nutrient transport kinetics will manifest themselves in potentially important differences in competitive advantage in modelled populations . Such differences become more apparent when considering the form of the relationship between nutrient status and Tmax ( Fig 7 ) , a topic that is also of consequence when describing the ammonium-nitrate interaction [71] . It is also important to couple nutrient-light limitations in the correct way , else the expected decrease in KG with light limitation does not occur [72] . Interactions with temperature and allometry are also complex [53 , 73] , with changes in cell size , overall growth rate , and differential impacts on transport vs metabolism [28 , 74] . All of this speaks to the importance of describing the relationship between multi-factor feedback interactions upon cell growth , with some attempt to simulate ( de ) repression of different metabolic pathways . In general , the importance and usefulness of using a single proxy as a determinant of competitive advantage seems overstated . This applies to usage of the value of kcat/KM in enzyme kinetics , Umax/KU in studies of diffusion limitation , or Gmax/KG in whole organism growth kinetics . Similarly , only considering stoichiometry represents too great a simplification in considerations of nutrient competition [59] . We simply have too limited knowledge of the real nutrient concentrations at the scales of consequence for these organisms ( proximate to the cells ) , while we also know that factors such as alternative nutritional routes ( nitrate vs ammonium vs dissolved organic -N; phosphate vs dissolved organic -P , phago-mixotrophy ) , different transporter types with different affinities for a given nutrient [14 , 16] , allelopathy [75] , palatability for grazers [76] and resistance to non-predator factors affecting cell mortality [77] are all important if not critical factors affecting competition at different times and places in the real world . Our analysis , like many other studies , makes the unrealistic caveat of all-else-being-equal across a wide range of organism types , shapes , sizes , motilities and stoichiometries . So , while Fig 4 portrays a general theoretical pattern , application of that pattern to explain species competition for growth in the same water body must be viewed with extreme caution . It is of some comfort that the approach justifies ( is consistent with ) a common assumption that fast growing ( r-select ) species are disadvantaged in mature ecosystems where their slower growing ( K-select ) competitors have a better nutrient affinity ( lower KG ) . However , simply relating KG ( or indeed any such parameter ) to size is in any case highly problematic: many very small , non-motile cells tend to grow together ( notably when P-stressed ) , and diatoms can grow in chains or mats , so that effective particle size ( affecting boundary layer thickness and sedimentation ) is often larger than it appears; the impacts of such changes are typically not included in models . Furthermore , little is known about interactions with alternative modes of nutrition , such as mixotrophy ( including the use of dissolved organics ) , which likely vary significantly between organisms and will impact greatly upon the significance of KG for a given limiting nutrient at any instant in time . Within simple bottom-up controlled systems operating under non-steady-state conditions , possession of a higher growth rate is expected to endow a powerful competitive advantage under conditions of nutrient abundance . Larger growing cells need not be disadvantaged in such systems . However , smaller organisms appear always to be at an advantage for nutrient acquisition within nutrient limited systems running closer to steady-state , as epitomised by chemostat experimental systems and typically observed in the oligotrophic oceans . In a chemostat , at a given dilution rate the substrate concentration converges on that which enables the growth rate to match the dilution rate . Besides the logistic challenges in running chemostats to determine KG , there is also the real risk that the organisms adapt to enforced slow growth over many months [66] . It is notable that the predicted values of KG from this study ( Fig 2 ) are in the main very low , bordering on the level of chemical detection in the bulk media , even when assuming a transporter protein nutrient affinity ( KT ) of 1 μM . Interestingly , in modelled systems , the dynamics of the system may be more heavily controlled by the parameters controlling activity of zooplankton than by the value of KG for phytoplankton [78] . It is also noteworthy that factors affecting cell size , motility/sedimentation , stoichiometry and cellular carbon density impact greatly upon predation kinetics and the value of the organism as food for the predator [79 , 26] . Thus , while motility enhances transport potential through decreasing boundary-layer limitations , motility is rather a double-edged-sword as it raises the likelihood of encountering a predator . For sure , simple comparisons between single-factors such as nutrient competition cannot possibly determine the true competitive advantage . We can perhaps be more secure in considering the implications of our analysis for the evolution of an individual species , where intra-species competition is important . Here , within a particular cell line of a given species , the values of KG and Gmax can be expected to be linked; a faster growing cell will have a higher KG . This observation is consistent with a general feature of enzyme activity such that high kcat is often associated with a high KM [42] , in consequence of a low KM being deleterious for the rapid breakdown of the enzyme-substrate complex . Irrespective of species-species interactions , one may thus expect a trade-off between KG and Gmax and for this to be reflected in the evolution of a particular cell line . Taken alone , this is an important trade-off between traits affecting the benefit of fast growth and is consistent with the observation that cells forced to grow slowly in a low-dilution chemostat ( noting that dilution rate = growth rate at steady-state , and that the residual nutrient concentration is lower at low dilution rates ) evolve a lower Gmax than the parent population [66] . The complexity of trade-offs in the evolution of individual enzymes [42] perhaps warns against attempting too-tight a linkage between KG and Gmax in terms of trait trade-off arguments at the whole organism level .
In the following we assume that the transporter rate density ( TRD ) has a maximum possible value ( TRDmax ) ; that is to say that , the plasma-membrane can only contain so-many nutrient transporter proteins over a given area . We assume TRDmax to be the experimentally determined maximum rate of 0 . 4 pgNμm-2 d-1 ( from the diatom Thalassiosira , using experimentally computed C-cell; Table 2 ) . Note , that the actual expressed value of TRD , and the instantaneous operation of transporter proteins , may be down-regulated due to long or short-term feedback linked to satiation and cellular nutrient status . It is assumed that all transporter proteins , contributing to TRD , have the same transport potential irrespective of the organism; hence we assume no features of the plasma-membrane or allied cell wall structure affect the functional value of kcat or KT of the embedded transporter proteins . The value of Tmax varies with the physiological status of the cells . Here we consider the N-status as indexed by the cellular N:C . The N-status is described as a normalised N:C quota [57] such that minimum stress is given by NCu = 1 , and maximum stress by NCu = 0 . The equation defining NCu is: NCu= ( 1+KQ ) ∙ ( NC−NCmin ) ( NC−NCmin ) +KQ∙ ( NCmax−NCmin ) ( 4 ) NC is the current cellular mass ratio of N:C , which ranges between NCmin when G is limited to 0 by supply of nutrient-N , and NCmax when G = Gmax . KQ is a curve shaping constant , which at a KQ = 10 gives the expected near-linear relationship between N:C and the growth rate , G [6] . The value of Tmax can be derived experimentally ( as in S2 Fig ) . Tmax can alternatively be described hypothetically as increasing with decreasing nutrient status . To achieve the latter , here we use a simple curve form that carries a minimum of Gmax × NCmax and rises rapidly as the N-status , NCu , decreases ( i . e . as N:C decreases from NCmax to NCmin ) . This equation contains a normalised RHt2 description which for values of 0≤NCu≤1 will return a value of 0 to 1 irrespective of the value of KTcon , which is a curve setting constant ( the lower the value the steeper the curve , increasing Tmax as N:C decreases with N-stress ) . Tmax=Gmax∙NCmax∙ ( 1+Tadd∙ ( 1+KTcon ) ∙ ( 1−NCu ) ( 1−NCu+KTcon ) ) ( 5 ) The value of Tadd provides a simple approach to reflect the diversity in scaling between the very highest expressed Tmax and that required to support G = Gmax , as broadly seen in real organisms ( S2 Fig ) . Tadd acts as a multiplier for Tmax ( dimensionless ) ; e . g . Tadd = 1 , will at NCu = 0 double the value of Tmax over that expressed when NCu = 1 with G = Gmax . If Tadd = 0 , then Eq 5 describes a flat Tmax , as is de facto assumed in most models [72 , 80] . The maximum possible value of Tadd in Eq 5 is a function of the value of TRDmax permitting us to explore the allometric and allied scaling of transport potential by reference to the maximum possible TRD ( which here we consider as 0 . 4 pg nutrient-N μm-2 d-1 ) and also to the value of TRD required to support Gmax , TRDGmax . From Eq 2 , we obtain: TRDGmax=Gmax∙NCmax∙CcellSA ( 6 ) Ccell is the C content per cell ( pgC ) ; this value as a function of ESD is described as per [9] . SA is the cell surface area ( μm2 ) , and NCmax is the mass N:C at G = Gmax . Tadd is then given by Tadd=TRDmax−TRDGmaxTRDGmax ( 7 ) From Fig 8 , it can be seen how the value of TRDGmax varies between organism configurations , increasing with size and Gmax . In particular large protists with their high demands for nutrients become limited by the value of TRDmax at high growth rates , i . e . TRDGmax approaches the maximum density of 0 . 4 pg nutrient-N μm-2 d-1 . Fig 6 , for a hypothetical organism with a fixed cellular carbon density ( C150 ) , shows the potential for smaller organisms to have scope for a far higher excess transport capacity; that is TRDmax: TRDGmax = δTRD is higher for small cells , and this excess is higher again at lower Gmax . However , in realty larger cells are less C-dense [9] , and this is even more apparent for diatoms as these are relatively even more vacuolated; this mitigates against the simple allometric response ( Fig 9; Cf . Fig 2 ) . From the value of Tmax , the transport rate ( T ) is given by Eq 8 ( Cf . Eq 1 ) , where S0 is the nutrient concentration at the plasma membrane surface , and KT is the half saturation constant for the nutrient transporter protein , T=Tmax∙S0S0+KT ( 8 ) This is rearranged to obtain S0: S0=TKTTmax−T ( 9 ) In reality , the value of T is limited by diffusion at low nutrient concentrations . This limitation sets a relationship between S∞ and S0 . From Eqs 16 and 17 in [25] , developed from [35] , the transport rate of nutrient into the cell ( T , ng cell-1 d-1 ) is related to the gradient between the bulk nutrient concentration and the nutrient concentration at the cell surface ( S∞−S0 , ng L-1 ) via the following equation: T=Dr ( 1+0 . 5∙rD∙c ) ∙4πr2 ( S∞−S0 ) ( 10 ) Here , r is the cell radius ( μm ) , D is diffusivity ( μm2 d-1 ) , c is the organism’s speed of motion either due to swimming or sedimentation ( μm d-1 ) . The thickness of the boundary layer impacts upon the difference between S∞ and S0; the larger the cell , and the slower its motion through the water , the greater is the value of ( S∞−S0 ) for a given value of T . By rearranging Eq 10 , we obtain the value of S∞ . Cell motility ( c , μm s-1 ) was configured using an empirical allometric equation using data from Sommer 1988 [81] and Visser & Kiørboe 2006 [82] according to [79] as: c=38 . 542* ( ESD ) 0 . 5424 ( 12 ) Sedimentation rates ( csed; μm s-1 ) were computed using Stoke’s law , from the cell radius ( r; μm ) , cell density ( ρorg; assumed here to be 1 . 0634 kg L-1 ) , seawater density ( ρw; assumed here to be 1 . 033 kg L-1 ) , dynamic viscosity ( η; assumed here to be 1 . 0846x 10−3 Pa s ) , and acceleration due to gravity ( g; 9 . 8 m s-2 ) . In order to compute the value of the bulk-water nutrient concentration ( S∞ ) that supports a given growth rate , the above equations were constructed to enable organism size , allometric parameters and motility to be altered . For given values of Gmax , NCmax and NCmin , the rate of N transport required to support a given G is computed . For a given cell size , cellular carbon density and N:C , we calculate the cell surface area , and the N-cell density at a given G . From these the rate of N-source transport per cell surface area is computed to support the given G; this is the value of T in Eqs 8 and 10 . | Relating environmental nutrient concentration and nutrient acquisition to cell growth is an important feature of numerical simulations describing ecological systems of microbes . Here we investigate the critical role of the combined effects of maximum growth rate , cell size , motion , and elemental stoichiometry on nutrient transport kinetics and thence growth kinetics . By applying mechanistic scaling of nutrient uptake our results identify fundamental shortcomings in the interpretation of empirically derived relationships used to describe nutrient uptake in microbes . While the amount of nutrient required to grow at a given rate under nutrient limited conditions increases rapidly with cell size , the maximum growth rate scales directly with the environmental nutrient concentration . Requiring less nutrient at lower maximum growth rates , cells can therefore remain healthier at lower resource abundance . Further , decreased carbon content per cell lowers demand for nutrient transport per surface area significantly . This allows larger phytoplankton cells , like diatoms , to significantly increase their competitive advantage with increasing sedimentation rates . These findings have important implications for numerical models both in a context of theoretical ecology and applied science . Our results highlight the importance of accounting for organism physiology and related feedbacks in ecological applications and climate change studies . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
]
| [
"invertebrates",
"cell",
"motility",
"medicine",
"and",
"health",
"sciences",
"phytoplankton",
"chemical",
"compounds",
"animals",
"nitrates",
"nutrition",
"materials",
"science",
"materials",
"physics",
"sedimentation",
"plants",
"chemistry",
"algae",
"physics",
"stoichiometry",
"eukaryota",
"cell",
"biology",
"diatoms",
"plankton",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"nutrients",
"protists",
"organisms"
]
| 2018 | Effects of growth rate, cell size, motion, and elemental stoichiometry on nutrient transport kinetics |
Infection with Listeria monocytogenes strains that enter the host cell cytosol leads to a robust cytotoxic T cell response resulting in long-lived cell-mediated immunity ( CMI ) . Upon entry into the cytosol , L . monocytogenes secretes cyclic diadenosine monophosphate ( c-di-AMP ) which activates the innate immune sensor STING leading to the expression of IFN-β and co-regulated genes . In this study , we examined the role of STING in the development of protective CMI to L . monocytogenes . Mice deficient for STING or its downstream effector IRF3 restricted a secondary lethal challenge with L . monocytogenes and exhibited enhanced immunity that was MyD88-independent . Conversely , enhancing STING activation during immunization by co-administration of c-di-AMP or by infection with a L . monocytogenes mutant that secretes elevated levels of c-di-AMP resulted in decreased protective immunity that was largely dependent on the type I interferon receptor . These data suggest that L . monocytogenes activation of STING downregulates CMI by induction of type I interferon .
Cell-mediated immunity ( CMI ) is a critical component for protection against intracellular pathogens . Upon infection , the innate immune response provides resistance and initiates the development of antigen-specific lymphocytes including cytotoxic CD8+ T cells , which ultimately kill host cells harboring pathogens [1] . The Gram-positive bacterium Listeria monocytogenes has been used for decades as a model organism to investigate the generation of CMI , as infection induces a robust effector and memory CD8+ T cell response that restricts bacterial growth following a lethal secondary challenge , resulting in long-lived sterilizing immunity [2] . Although it is generally agreed that activation of the innate immune system is critical for the initiation of adaptive immunity [3] , the specific signaling pathways necessary to elicit a robust protective immune response to L . monocytogenes remain poorly understood . L . monocytogenes is detected by multiple innate immune signaling pathways during infection [4] . Following engulfment by macrophages and dendritic cells , the bacteria reside within phagosomes where they are detected by Toll-Like Receptors ( TLRs ) , resulting in the activation of MyD88-dependent response genes [5] . By secreting a pore-forming cytolysin , listerolysin O ( LLO ) , L . monocytogenes escapes into the cytosol where it replicates and polymerizes actin to facilitate cell-to-cell spread [6] . L . monocytogenes is detected by several cytosolic innate immune pathways leading to a cytokine profile distinct from that of LLO-deficient bacteria , which are restricted to the phagosome [5] , [7] . The primary cytosolic sensor of L . monocytogenes is STING ( stimulator of interferon ( IFN ) genes , also known as MPYS , MITA and ERIS ) , an ER-localized transmembrane protein [8] . STING is activated by cyclic dinucleotides ( CDNs ) that are either produced by a pathogen or by an endogenous cyclic GMP-AMP synthase that is activated by DNA [9] , [10] . Direct binding of CDNs to STING activates a downstream signaling cascade involving TBK1 and IRF3 [11] , [12] , [13] . In the case of L . monocytogenes , cyclic diadenosine monophosphate ( c-di-AMP ) is secreted through bacterial multi-drug efflux pumps , leading to STING activation and transcription of IFN-β and co-regulated genes [14] , [15] . STING-deficient macrophages or mice are unable to produce IFN-β in response to L . monocytogenes infection indicating that STING is required for the type I IFN response to L . monocytogenes [16] , [17] . Purified CDNs are immunostimulatory in vitro and in vivo . Murine and human dendritic cells exposed to cyclic diguanosine monophosphate ( c-di-GMP ) or c-di-AMP exhibit enhanced surface expression of costimulatory markers and T cell proliferation . Mice mount a significant antibody response following co-administration of protein antigens with c-di-GMP or c-di-AMP [13] , [18] , [19] , [20] . CDNs also stimulate cellular immune responses . Antigen-stimulated splenocytes from mice immunized with β-ααgalactosidase in the presence of c-di-GMP or c-di-AMP proliferate and secrete cytokines [18] , [19] . These data indicate that CDNs are sufficient to elicit a cell-mediated adaptive immune response . Entry of L . monocytogenes into the host cytosol is necessary to generate secondary protective immunity , as phagosome-restricted heat-killed or LLO-deficient bacteria do not elicit functional cytotoxic T cells and long-term memory responses [21] , [22] , [23] . The attenuated ActA-deficient mutant strain , which escapes the phagosome but fails to polymerize actin and spread to neighboring cells , is fully immunogenic to mice [24] . Furthermore , MyD88-deficient mice , while highly susceptible to acute infection with virulent L . monocytogenes , are fully protected following secondary lethal challenge when immunized with the ActA-deficient mutant [25] , [26] , [27] , [28] . These findings suggest that phagosomal detection of L . monocytogenes during immunization is not sufficient for the development of protective immunity . STING activation induces an array of IRF3-dependent genes [5] as well as NF-κB and STAT6-dependent genes [29] , [30] . Since LLO-deficient bacteria fail to enter the cytosol and induce STING-related genes [5] , [7] , we hypothesized that the detection of L . monocytogenes by STING is required for CMI . In this study , we tested whether STING signaling plays an important role in the generation of protective immunity to L . monocytogenes .
In the model of protective immunity used in these studies , mice were immunized with an attenuated yet immunogenic strain of L . monocytogenes that lacks the actA and inlB genes ( ActA−Lm ) and challenged 30–38 days later with 2LD50 ( 2×105 colony forming units ( CFU ) ) of wild-type L . monocytogenes ( WT Lm ) . Previous studies typically immunize mice with 0 . 1LD50 of L . monocytogenes ( 1×107 CFU ActA−Lm for C57BL/6 mice ) [21] . At this high immunization dose , bacterial burdens in subsequently challenged mice are below the limit of detection . In contrast , a lower immunization dose of 103 CFU ( ∼0 . 00001LD50 for C57BL/6 mice ) still generated significant immunity as compared to naïve mice , but did not induce saturating immunity and thus revealed differences that might be missed using higher doses ( Fig . S1A ) . To determine whether STING signaling is required for the generation of protective immunity to L . monocytogenes , mice lacking STING ( Goldenticket , Gt ) were immunized with 103 CFU of ActA−Lm expressing ovalbumin ( ActA−Lm-OVA ) and challenged 30–38 days later with 2LD50s of WT Lm-OVA . Surprisingly , whereas naïve STING-deficient mice had similar bacterial burdens as naïve C57BL/6 ( B6 ) mice , immunized STING-deficient mice had approximately 1 . 5–2 logs fewer bacteria in spleens and livers compared to immunized B6 mice ( Fig . 1A ) . At higher immunization doses ( 104 and 105 CFU ) , the majority of B6 and STING-deficient mice had bacterial numbers below the limit of detection in the spleen and thus no significant differences could be observed ( Fig . S1B ) . Since cytotoxic CD8+ T cells are the major mediator of L . monocytogenes clearance following secondary challenge [2] , [31] , the number of OVA-specific CD8+ T cells were measured by staining splenocytes with a MHC class I restricted OVA tetramer ( Kb/OVA257–264 ) . STING-deficient mice had significantly higher total numbers of OVA-specific CD8+ T cells compared to B6 mice ( Fig . 1B ) . These data suggested that STING-deficient mice exhibited enhanced immunity . STING stimulation leads to the activation of the transcription factor , IRF3 [12] . In addition , IRF7 contributes to IFN-α production in response to L . monocytogenes in vivo [32] . To determine the role of these downstream effectors of STING signaling , mice lacking IRF3 ( IRF3−/− ) or IRF3 and IRF7 ( IRF3/7−/− ) were examined for protective immunity . Compared to B6 mice , IRF3-deficient mice had less bacteria in the livers and IRF3/7-deficient mice had less bacteria in the spleens and livers ( Fig . 1A ) . Both groups had significantly higher numbers of OVA-specific CD8+ T cells than B6 mice ( Fig . 1B ) , suggesting that these transcription factors contribute to the STING-mediated decrease in immunity . To evaluate whether T cells from mice lacking STING signaling possessed effector functions , splenocytes from ActA−Lm-OVA-immunized mice were stimulated with either the MHC class I-restricted peptide OVA257–264 , or the MHC class II-restricted peptide LLO190–201 from L . monocytogenes and measured for IFN-γ , TNF-α and IL-2 production by intracellular cytokine staining and flow cytometry . STING-deficient mice had significantly higher numbers of polyfunctional IFN-γ- , TNF-α- and IL-2- producing CD8+ and CD4+ T cells compared to B6 mice ( Fig . 1C ) . IRF3/7-deficient but not IRF3-deficient mice also had higher numbers of IFN-γ- and TNF-α-producing CD8+ T cells . These data suggested that in the absence of STING or its downstream transcription factors IRF3 and IRF7 , CD8+ T cell expansion , cytokine production and bacterial clearance was enhanced . L . monocytogenes activates TLRs which signal via the adaptor MyD88 [4] . It is possible that both MyD88- and STING-dependent response genes play redundant roles in generating protective immunity to L . monocytogenes . To test this hypothesis , we bred mice lacking both MyD88 and STING ( MyD88−/−Gt ) . Bone marrow-derived macrophages ( BMMs ) from MyD88/STING-deficient mice infected with WT Lm had low or non-detectable expression of the cytokines IFN-β , IL-12 p40 , TNF-α and IL-6 ( Fig . S2A ) . To examine whether the loss of both the MyD88 and STING signaling pathways affect bacterial clearance during an acute infection , B6 , MyD88- and MyD88/STING-deficient mice were infected with WT Lm . MyD88/STING-deficient mice had similar bacterial burdens as MyD88-deficient mice early after infection , however by day 3 , had significantly higher CFU in the spleen and the liver ( Fig . S2B ) . Although mice lacking MyD88 cannot survive infection with WT Lm , MyD88/STING-deficient mice died earlier than MyD88-deficient mice ( data not shown ) . These data indicated that in the absence of MyD88 , STING contributes to L . monocytogenes clearance during an acute response . To determine whether the loss of MyD88 and STING affects initiation of the adaptive response to L . monocytogenes , the upregulation of costimulatory molecules on splenic dendritic cells from immunized B6 , STING- , MyD88- and MyD88/STING-deficient mice was measured . Surface expression of CD86 and CD40 was decreased in MyD88- and STING-deficient mice and further reduced in the MyD88/STING-deficient mice , suggesting an additive effect of these two pathways ( Fig . 2A ) . Upregulation of the activation marker CD69 on CD8+ and CD4+ T cells was also decreased in both MyD88- and STING-deficient mice and ablated in MyD88/STING-deficient mice ( Fig . 2B ) . Furthermore , IL-6 , TNF-α and MCP-1 in the serum were mostly dependent on MyD88 but further reduced in the MyD88/STING-deficient mice . IL-12 p70 was non-detectable in either MyD88- or MyD88/STING-deficient mice ( Fig . 2C ) . These data indicated that mice lacking MyD88 and STING signaling had significantly reduced dendritic and T cell activation and cytokine production in vivo in response to L . monocytogenes immunization . Mice lacking both MyD88 and STING were tested for the ability to develop protective immunity . Similar to B6 and MyD88-deficient mice , MyD88/STING-deficient mice were protected following secondary lethal challenge and showed no signs of disease unlike naïve mice which were either moribund or dead at the time of sacrifice . MyD88/STING-deficient mice had a small increase in the number of bacteria in the livers compared to MyD88-deficient mice ( Fig . 3A ) . However , MyD88/STING-deficient mice had significantly higher number of OVA-specific CD8+ T cells compared to MyD88-deficient mice ( Fig . 3B ) , suggesting that the higher bacterial loads in the liver was likely due to loss of innate rather than adaptive immune responses . These data indicated that regardless of the presence or absence of MyD88 , protective immunity to L . monocytogenes is enhanced in the absence of STING . Mice immunized with LLO-deficient L . monocytogenes fail to develop protective immunity [21] , [22] . Since LLO-deficient bacteria do not activate STING , we hypothesized that c-di-AMP-mediated STING activation might be sufficient to restore immunity in LLO−Lm-immunized mice . In support of previous reports , we found that BMDCs stimulated with c-di-AMP secreted cytokines and upregulated costimulatory markers in vitro ( Fig . S3 ) . Furthermore , mice administered c-di-AMP intravenously exhibited increased surface expression of CD86 and CD40 on splenic dendritic cells and of the activation marker CD69 on splenic CD8+ and CD4+ T cells in a STING-dependent manner ( Fig . 4A and 4B ) , indicating that c-di-AMP can induce an inflammatory response in our model . To test our hypothesis , B6 mice were immunized with LLO−Lm-OVA in the presence of c-di-AMP . While ActA−Lm-OVA-immunized mice restricted bacterial growth following challenge , the co-administration of c-di-AMP with LLO−Lm-OVA did not rescue protective immunity ( Fig . 4C ) . Instead , the presence of c-di-AMP significantly reduced immunity in ActA−Lm-OVA-immunized mice suggesting that STING signaling inhibits protective immunity to L . monocytogenes infection . To further examine the effect of c-di-AMP on immunity to L . monocytogenes , B6 , STING- and IRF3/7-deficient mice were immunized with ActA−Lm-OVA in the presence or absence of c-di-AMP . Following challenge , B6 mice immunized in the presence of c-di-AMP had significantly higher bacterial numbers and had fewer numbers of total and OVA-specific CD8+ T cells in the spleen than mice immunized with ActA−Lm-OVA alone ( Fig . 4D , 4E and S4A ) . STING-deficient mice were protected and had a robust CD8+ T cell expansion confirming that c-di-AMP-mediated inhibition of immunity was STING-dependent . IRF3/7-deficient mice were significantly protected as compared to naïve mice and had a population of OVA-specific CD8+ T cells , suggesting that c-di-AMP-mediated inhibition is partially due to IRF3 and IRF7 ( Fig . 4D and 4E ) . Interestingly , IRF3/7-deficient mice were less protected compared to mice immunized with ActA−Lm-OVA alone , indicating that STING-dependent , IRF3/7-independent signaling also plays a role in loss of protective immunity . Next , we evaluated a L . monocytogenes mutant , tetR::Tn917 , that secretes 20-fold more c-di-AMP than WT L . monocytogenes [14] , [15] . B6 , STING- and IRF3/7-deficient mice were immunized with either ActA−Lm-OVA or the tetR::Tn917 mutant in the ActA−Lm-OVA background ( tetRActA−Lm-OVA ) . B6 mice immunized with tetRActA−Lm-OVA had significantly higher bacterial numbers in the spleen compared to ActA−Lm-OVA-immunized mice whereas STING- and IRF3/7-deficient mice exhibited no significant difference ( Fig . 5A ) . Furthermore , B6 mice immunized with tetRActA−Lm-OVA had a smaller population of total and OVA-specific CD8+ T cells compared to ActA−Lm-OVA-immunized mice ( Fig . 5B and S4B ) . These data supported our finding that enhanced STING signaling lead to a reduction in protective immunity . To determine whether enhanced STING signaling reduces T cell priming , OVA-specific CD8+ T cells were measured at the peak of the primary response from mice immunized with ActA−Lm-OVA in the presence or absence of c-di-AMP or tetRActA−Lm-OVA . At 7 days post immunization , mice immunized in the presence of c-di-AMP had significantly fewer OVA-specific CD8+ T cells compared to mice immunized with ActA−Lm-OVA alone . Mice immunized with the tetRActA−Lm-OVA mutant also had a small decrease in antigen-specific cells ( Fig . 6A ) . Furthermore , the small population of OVA-specific CD8+ T cells that were present in mice immunized with elevated STING activation had significantly higher surface expression of the naïve T cell maker CD62L , suggesting that enhanced STING signaling elicited fewer antigen-specific effector T cells ( Fig . 6B ) . In addition , peptide-stimulated CD8+ and CD4+ splenocytes from ActA−Lm-OVA-immunized mice in the presence of c-di-AMP or tetRActA−Lm-OVA-immunized mice produced fewer cytokines than those from ActA−Lm-OVA-immunized mice ( Fig . 6C ) . These data indicate that mice immunized in the presence of enhanced STING signaling exhibited reduced T cell priming . To determine the role of type I IFNs in STING-mediated CMI , IFN-αβR-deficient mice were tested for protective immunity . IFN-αβR-deficient mice restricted bacterial growth better than B6 mice ( Fig . 7A and 7B ) , indicating that like STING- and IRF3/7-deficient mice , IFN-αβR-deficient mice also hyper-immunize . Although IFN-αβR-deficient mice immunized in the presence of c-di-AMP had higher bacterial numbers compared to mice immunized with ActA−Lm-OVA alone , these mice were significantly more protected than naïve mice , indicating a role for both type I IFN-dependent and independent mechanisms of suppression ( Fig . 7A ) . IFN-αβR-deficient mice immunized with tetRActA−Lm-OVA were completely protected ( Fig . 7B ) . These data indicated that the c-di-AMP-mediated inhibition of protective immunity is largely dependent on type I IFNs . Interestingly , we found that mice immunized in the presence of the synthetic double-stranded RNA , polyinosinic∶polycytidylic acid ( poly ( I∶C ) ) , a STING-independent agonist of TLR3 and IRF3 , also lost the ability to restrict bacterial growth following challenge ( Fig S5 ) , suggesting that type I IFN-mediated inhibition of immunity is unlikely STING specific . We next determined whether inhibition of T cell priming by STING-dependent type I IFNs acted directly on lymphocytes . CD8+ T cells lacking the IFN-αβR undergo clonal expansion in response to primary L . monocytogenes infection so an adoptive transfer model could be used [33] , [34] . B6 , STING- , or IFN-αβR-deficient mice were injected with WT and IFN-αβR-deficient OT-I splenocytes and subsequently immunized with ActA−Lm-OVA in the presence or absence of c-di-AMP . At 7 days , WT and IFN-αβR-deficient OT-I cells expanded in ActA−Lm-OVA-immunized B6 mice , whereas in the presence of c-di-AMP , both WT and IFN-αβR-deficient OT-I cells had significantly reduced populations indicating that type I IFNs were not directly blocking T cell priming ( Fig . 8A and 8B ) . Inhibition of T cell expansion by c-di-AMP was rescued in IFN-αβR- and STING-deficient mice further indicating that type I IFN-mediated suppression of immunity is not T cell intrinsic .
The results of this study show that the STING signaling pathway is not required to elicit CMI to L . monocytogenes . In fact , the absence of STING or IRF3 and IRF7 led to a higher number of antigen-specific CD8+ T cells and increased levels of protective immunity . Mice lacking both MyD88 and STING were also protected upon secondary challenge , indicating that not only is STING dispensable for the generation of a protective response to L . monocytogenes , it does not act in a redundant fashion with the TLR-MyD88 signaling pathway . Conversely , when STING activity was enhanced either by administering c-di-AMP during immunization or using a bacterial mutant that secretes elevated levels of c-di-AMP , mice failed to immunize and had decreased numbers of antigen-specific T cells following reinfection . This suppressive effect was largely due to the induction of type I IFNs since IFN-αβR-deficient mice immunized with either ActA−Lm-OVA in the presence of c-di-AMP or the tetR mutant were protected . Collectively , these findings suggest that L . monocytogenes-induced STING activation reduces the host adaptive immune response by induction of type I IFN . The mechanism of type I IFN-mediated inhibition of T cell priming remains unclear . We found that both WT and IFN-αβR-deficient CD8+ T cells had significantly reduced expansion in the presence of c-di-AMP , suggesting that c-di-AMP-mediated suppression of lymphocyte priming is T cell extrinsic . One possibility is that the uptake of type I IFN-induced apoptotic cells by macrophages results in the release of the immunosuppressive cytokine IL-10 [35] . However , mice administered c-di-AMP did not have detectable levels of IL-10 in the serum . Furthermore , similar to B6 mice , IL-10-deficient mice immunized in the presence of c-di-AMP were unable to elicit OVA-specific CD8+ T cells following challenge ( data not shown ) . Thus , IL-10 is not the downstream effector of STING/type I IFN-mediated suppression of adaptive immunity in our model of protective immunity . Other mechanisms including indoleamine 2 , 3-dioxygenase ( IDO ) -mediated T cell suppression , which is upregulated by type I IFN [36] , may play a role in c-di-AMP-mediated inhibition of immunity . Indeed , Huang et al . found that splenic DCs from mice treated with DNA nanoparticles suppress ex vivo T cell proliferation in a STING- and IDO-dependent manner [37] . While spleens from immunized STING-deficient mice contained fewer bacteria than B6 mice , bacterial clearance by immunized IRF3/7- and IFN-αβR-deficient mice was even higher . One possibility is that there are STING-independent sources of type I IFNs in response to L . monocytogenes . Supporting this hypothesis , we observed that MyD88/STING-deficient macrophages produced low levels of IFN-β eight hours post infection . Recent studies found that RIG-I-deficient cells had reduced IFN-β secretion in response to L . monocytogenes infection suggesting that RNA from L . monocytogenes can be detected by the host [38] , [39] . Supporting this hypothesis , we found that immunized mice lacking MAVS , the signaling adaptor for RIG-I , were more protected than control mice upon reinfection ( data not shown ) . Thus , RIG-I-dependent , STING-independent induction of type I IFNs via IRF3 and IRF7 may also contribute to the inhibition of protective immunity . Another possibility is that since naïve IRF3/7- and IFN-αβR-deficient mice exhibit heightened bacterial clearance compared to both naïve B6 and STING-deficient mice , innate immune factors are also likely restricting reinfection . In fact , IFN-αβR-deficient mice have higher neutrophil recruitment in response to L . monocytogenes infection [40] . Thus , careful examination of each mouse strain will be necessary to determine the extent of the contribution of innate versus adaptive immune mechanisms in protective immunity . Type I IFN-mediated inhibition of immunity is unlikely specific to STING activation . Mice immunized in the presence of the poly ( I∶C ) , which induces type I IFNs independently of STING , also lost the ability to restrict bacterial growth following challenge . Several studies have shown that administering poly ( I∶C ) prior to a protein antigen inhibits clonal expansion of antigen-specific CD8+ T cells , supporting our findings that systemic type I IFN-induced inflammation reduces T cell priming [41] , [42] . The importance of type I IFNs during bacterial infection is less understood than for viruses [43] . For example , IFN-β is the highest upregulated gene following L . monocytogenes cytosolic invasion , yet mice lacking the IFN-αβR are more resistant to acute infection , suggesting that type I IFNs may promote pathogenesis [35] , [44] , [45] . However , strains of L . monocytogenes that secrete elevated levels of c-di-AMP and induce higher levels of type I IFNs , are not hypervirulent [46] , but induce considerably less T cell immunity as shown in this study . Thus it is possible that secretion of c-di-AMP and consequent type I IFN production may play a role in L . monocytogenes pathogenesis by suppressing the development of adaptive immunity . Although L . monocytogenes generally causes acute infections , recent studies have found that type I IFNs promote chronic infections with LCMV by suppressing cell-mediated mechanisms of viral control [47] , [48] . In the case of Mycobacterium tuberculosis , mice lacking IRF3 are more resistant to infection with M . tuberculosis suggesting that IRF3 activation is detrimental to host clearance [49] . In humans , IFN-α treatment leads to higher incidences of TB reactivation [50] . Furthermore , active TB patients exhibit an increase in type I IFN-inducible transcripts in the blood , which correlated with disease severity [51] . Therefore , type I IFNs may exacerbate or maintain secondary or long-term chronic infections . Interestingly , human STING often contains polymorphisms that makes it resistant to bacterial but not host derived CDNs [52] . STING-mediated suppression of protective immunity was not solely due to type I IFNs . Although protective immunity in tetRActA−Lm-OVA-immunized mice was rescued in the absence of IRF3/7 and the IFN-αβR , mice immunized in the presence of c-di-AMP exhibited type I IFN-independent suppression . Since STING activates NF-κB as well as IRF3 , it is possible that NF-κB-dependent inflammation also plays a role in restricting immunity . We believe that administering c-di-AMP activates STING more robustly compared to infection with the tetRActA−Lm strain and thus inhibition of immunity by type I IFN-independent inflammation would become more apparent . The results of this and other studies suggest an inverse relationship between the extent of inflammation and the development of adaptive immunity . For example , IL-12-deficient mice immunized with L . monocytogenes develop higher numbers of CD8+ memory T cells and are more resistant to reinfection [53] . In previous studies , we found that a L . monocytogenes strain engineered to activate the inflammasome , and consequently induce high levels of IL-1β secretion , was a poor inducer of adaptive immunity [54] . Furthermore , co-administration of heat-killed or LLO-deficient L . monocytogenes blocked immunity to WT bacteria in a MyD88-dependent manner [55] . Thus , activation of three distinct signaling pathways , STING , MyD88 , and caspase-1 all resulted in the inhibition of the development of adaptive immunity . Therefore , there appears to be a dichotomy between innate immune pathways that are necessary for survival ( for example , MyD88 for L . monocytogenes and type I IFN for viruses ) , and those that lead to adaptive immunity . In fact , our data and work from others suggest that lack of inflammation represents an ideal environment for the generation of memory T cells [31] , [56] . Indeed , mice deficient for both MyD88 and STING are fully immunized by L . monocytogenes even though there was a significant reduction of dendritic and T cell activation and cytokine production following immunization . Considering that the innate immune response is believed to be required for the initiation of adaptive immunity , we were surprised that MyD88/STING-deficient mice immunized with L . monocytogenes were protected after reinfection . This raises the question , which innate immune signaling pathways contribute to the initiation of T cell priming to L . monocytogenes ? Previous work from our group found that NOD2 detects cytosolic L . monocytogenes [5] . However , immunized MyD88/NOD1/2-deficient mice clear bacteria upon secondary lethal challenge ( data not shown ) , suggesting that protective immunity is not due to redundancy between MyD88 and NOD-like signaling pathways . Future studies to identify which innate immune detection pathways are required for L . monocytogenes-mediated CMI would provide a greater understanding of how pathogens and adjuvants elicit protective immunity , knowledge that can be used for the development and improvement of vaccines .
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 . All protocols were reviewed and approved by the Animal Care and Use Committee at the University of California , Berkeley ( MAUP# R235-0813B ) . C57BL/6 mice were purchased from The Jackson Laboratory . Goldenticket ( Gt ) mice were generated from an ENU mutagenesis screen . Gt mice contain a single nucleotide mutation in STING resulting in the absence of the STING protein [17] . All mice were in the C57BL/6 genetic background . Gt , MyD88−/− , MyD88−/−Gt , IRF3−/− , IRF3/7−/− and IFNAR1−/− mice were bred in our facilities . GFP+/+ OT-I+/+ RAG2−/− and IFNAR1−/− OT-I+/+ Ly5 . 1+/+ RAG2−/− mice were generously provided by Ellen Robey . All L . monocytogenes strains were in the 10403S background . ActA−Lm-OVA ( ΔactAΔinlB ) ( DP-L6014 ) [57] , WT Lm-OVA ( DP-L6018 ) and LLO−Lm-OVA ( Δhly ) ( DP-L6017 ) [21] L . monocytogenes were previously described . For tetRActA−Lm-OVA ( DP-L6015 ) , the tetR::Tn917 transposon [14] was transduced into ActA−Lm-OVA . L . monocytogenes were grown in brain heart infusion ( BHI ) media at 30°C overnight without shaking to stationary phase . For in vitro infections , L . monocytogenes was washed 3× in PBS . For in vivo infections , L . monocytogenes was diluted in BHI at 37°C shaking for ∼2 hours until they reached an OD600 0 . 4–0 . 6 . Eight to 12 week old sex-matched mice were infected intravenously with 103 CFU ( unless otherwise indicated ) of L . monocytogenes diluted in phosphate buffered saline ( PBS ) in a total volume of 200 µl . For acute infections and primary immunization studies , mice were sacrificed at 1 , 2 and 3 or 7 days post infection , respectively . For challenge studies , mice immunized 30–38 days prior were infected with 2×105 CFU of WT Lm-OVA . Where indicated , mice were administered either 50 µg or 100 µg of c-di-AMP or 50 µg of poly ( I∶C ) ( InvivoGen ) with the bacterial inoculum . Three days later , spleens and livers were homogenized in 0 . 1% IGEPAL CA-630 ( Sigma ) and plated on LB-strep plates to enumerate CFU . For analysis of CD8+ T cell responses , spleens were divided and weighed . For splenic dendritic and T cell activation , mice were immunized with 105 CFU of ActA−Lm-OVA . A higher dose was used to allow for the easy detection of activated cells . For OT-I cells , splenocytes from GFP+/+ OT-I+/+ RAG2−/− and IFNAR1−/− OT-I+/+ Ly5 . 1+/+ RAG2−/− mice were isolated and washed 3× with PBS . Percent of CD8+ OT-I cells was determined by flow cytometry . 2×104 of each cell type was injected intravenously into each mouse ( 4×104 total cells/mouse ) 1 day prior to immunization with L . monocytogenes . C-di-AMP was generated by Josh Woodward as previously described [58] . Purified c-di-AMP was resuspended in tissue culture grade PBS . LPS was removed from the prepared c-di-AMP using Detoxi-gel endotoxin removing gel ( Pierce ) according to the manufacturers instructions . Endotoxin content was measured using the Toxinsensor Chromogenic LAL Endotoxin assay kit ( Genescript ) . The nucleotide solution was passed through the Detoxi-gel until endotoxin levels were <0 . 0125 EU/ml . Nucleotide was then diluted to 500 µg/ml . For analysis of T cell responses , spleen halves were dissociated and filtered through a 70 µm cell strainer . Red blood cells were lysed with Red Blood Cell Lysing Buffer ( Sigma ) . To determine OVA-specific cells , splenocytes were stained with anti-mouse CD8 , CD44 , CD62L and a Kb/OVA257–264 tetramer . Representative FACs plots are gated on CD8+ cells and values show the median percentage of Kb/OVA257–264 tetramer+ CD44+ within the CD8+ cell population ± SEM . For peptide stimulation assays , splenocytes were stimulated for 5 hours with 2 µM OVA257–264 or LLO190–201 peptide in the presence of GolgiPlug ( BD Biosciences ) . Cells were surface stained with anti-mouse CD8 and CD4 , fixed and permeabilized using Cytofix/Cytoperm ( BD Biosciences ) , and stained for intracellular anti-mouse IFN-γ , TNF-α and IL-2 . For splenic dendritic cells and T cells , splenocytes were stained with anti-mouse CD11b , CD11c , CD86 and CD40 or anti-mouse CD8 , CD4 and CD69 , respectively . Flurophore-conjugated antibodies were purchased from eBioscience . Samples were acquired using an LSRII flow cytometer ( BD Biosciences ) and analyzed using FlowJo software ( Tree Star ) . BMMs were generated as previously described [59] . In a 6-well plate , 2×106 BMMs were either infected with WT Lm at a multiplicity of infection ( MOI ) of 2 bacteria per cell or stimulated with 10 µM c-di-AMP . At 30 minutes post infection , gentamicin was added for a final concentration of 50 µg/ml . At 4 and 8 hours post infection , cells were harvested and RNA was purified using the RNAqueous kit ( Ambion ) . RNA was then DNase treated , processed and analyzed as previously described [5] . Bone marrow from femurs was plated in media containing 20 ng/ml recombinant murine GMCSF ( ProSpec ) at a density of 5×105 cells/ml in a 24-well plate . At days 2 , 4 and 5 media was replaced with fresh media containing 20 ng/ml GMCSF and cells were harvested on day 6 . BMDCs were plated at 3×105 cells/well in 48-well plates . Cells were incubated with either 10 µM c-di-AMP , 100 ng/ml lipopolysaccharide ( LPS ) , or 20 µg/ml poly ( I∶C ) ( InvivoGen ) . After 24 hours , supernatant was assayed for IFN-β using ISRE-L929 cells as previously described [14] , or for MCP-1 , IL-12p40 ( BD OptEIA kit , BD Biosciences ) and IL-6 ( eBioscience ) by ELISA . Serum cytokines were measured using the CBA Mouse Inflammation Kit ( BD Biosciences ) and analyzed on the LSRII flow cytometer . A two-tailed , Mann-Whitney U test was used to analyze the significance of differences in the means between groups . Significance is indicated as * p<0 . 05 , ** p<0 . 005 , *** p<0 . 0005 or ns = not significant . | Current vaccines are successful at generating neutralizing antibodies , however there is a pressing medical need to find adjuvants that yield long-lived memory T cells . Immunization with the bacterium Listeria monocytogenes induces a robust protective immune response mediated by cytotoxic lymphocytes that are efficient at killing infected cells upon reinfection . When L . monocytogenes enters a cell , it secretes the small molecule cyclic diadenosine monophosphate ( c-di-AMP ) , which activates the host protein STING leading to a type I interferon response . In this study , we tested whether STING activation plays a role in the generation of cytotoxic lymphocytes and protective immunity using a mouse immunization model . We found that in the absence of STING signaling mice restricted bacterial growth and maintained higher numbers of cytotoxic lymphocytes upon reinfection , whereas mice immunized in the presence of elevated levels of c-di-AMP were less protected . These results suggest that the inflammation induced by a bacterial pathogen can be detrimental to the development of adaptive immunity , which could provide new insights into vaccine development . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| []
| 2014 | STING-Dependent Type I IFN Production Inhibits Cell-Mediated Immunity to Listeria monocytogenes |
Iron plays an essential role in many biological processes , but also catalyzes the formation of reactive oxygen species ( ROS ) , which can cause molecular damage . Iron homeostasis is therefore a critical determinant of fitness . In Caenorhabditis elegans , insulin/IGF-1 signaling ( IIS ) promotes growth and reproduction but limits stress resistance and lifespan through inactivation of the DAF-16/FoxO transcription factor ( TF ) . We report that long-lived daf-2 insulin/IGF-1 receptor mutants show a daf-16–dependent increase in expression of ftn-1 , which encodes the iron storage protein H-ferritin . To better understand the regulation of iron homeostasis , we performed a TF–limited genetic screen for factors influencing ftn-1 gene expression . The screen identified the heat-shock TF hsf-1 , the MAD bHLH TF mdl-1 , and the putative histone acetyl transferase ada-2 as activators of ftn-1 expression . It also revealed that the HIFα homolog hif-1 and its binding partner aha-1 ( HIFβ ) are potent repressors of ftn-1 expression . ftn-1 expression is induced by exposure to iron , and we found that hif-1 was required for this induction . In addition , we found that the prolyl hydroxylase EGL-9 , which represses HIF-1 via the von Hippel-Lindau tumor suppressor VHL-1 , can also act antagonistically to VHL-1 in regulating ftn-1 . This suggests a novel mechanism for HIF target gene regulation by these evolutionarily conserved and clinically important hydroxylases . Our findings imply that the IIS and HIF pathways act together to regulate iron homeostasis in C . elegans . We suggest that IIS/DAF-16 regulation of ftn-1 modulates a trade-off between growth and stress resistance , as elevated iron availability supports growth but also increases ROS production .
In order to survive in a changing environment , organisms have evolved abilities to sense their surroundings and adaptively adjust their physiology . For example , the nematode Caenorhabditis elegans is capable of postponing reproduction if conditions are unsuitable for growth and reproduction by forming dauer larvae [1] , [2] , [3] . This developmentally arrested third larval stage is resistant to starvation and other stressors , allowing the animal to survive until conditions improve . Should this occur , dauer larvae can re-enter the normal reproductive life cycle . The decision between reproductive growth and survival with enhanced stress resistance is controlled by a complex sensory/signaling network that includes the insulin/IGF-1 signaling ( IIS ) pathway [2] . Mutants with reduced IIS exhibit constitutive dauer larva formation , but can also form adults that are resistant to a range of stressors , including reactive oxygen species ( ROS ) , UV irradiation , heat stress and ER stress [4] , [5] , [6] . IIS controls this response through the DAF-16/FoxO transcription factor , which enters the nucleus under adverse conditions and affects gene regulation [7] , [8] . DAF-16 promotes increased expression of many genes encoding proteins that protect against stress , including superoxide dismutases , drug metabolizing enzymes and molecular chaperones 9 , 10 . DAF-16 is also required for the longevity of IIS mutants , for example those with defects in the DAF-2 insulin/IGF-1 receptor [11] . Both of these roles of DAF-16 , the promotion of stress resistance and longevity , will improve the chances of living through periods of adversity . Whether the same downstream mechanisms cause increased stress protection and longevity remains unclear [12] . One factor contributing to growth and stress resistance is cellular iron availability . Free intracellular iron is toxic to the cell due to its role in catalyzing the Fenton reaction , which generates hydroxyl radicals from hydrogen peroxide:However , while free intracellular iron is harmful to the cell , iron is also an important element for a large number of cellular processes , including electron transport , deoxyribonucleotide synthesis , cellular detoxification , the cell cycle , oxygen transport and many others [13] , [14] . Lack of iron is thought to affect the health of up to a billion people worldwide [15] . As well as nutritional iron deficiency , disruption of mechanisms that regulate iron homeostasis can also lead to a number of serious diseases in humans , such as anemias and iron overload disorders [16] , [17] . The maintenance of appropriate iron levels is therefore important to viability and is tightly regulated by a number of proteins . These include ferritins , which form 24-subunit spherical nanocages that are each able to safely store up to 4500 atoms of iron . Heavy chain ferritins ( H-ferritins ) contain a ferroxidase centre , which has the capacity to convert Fe ( II ) to Fe ( III ) when the iron atom enters the complex [18] . The C . elegans genome contains two H-ferritin genes , ftn-1 and ftn-2 [19] . ftn-1 is predominantly expressed in the intestine , while ftn-2 is expressed in many cell types [19] , [20] . In vertebrates , regulation of ferritin gene expression in response to iron levels is achieved both transcriptionally [21] , and post-transcriptionally by the actions of iron regulatory proteins ( IRPs ) which bind to iron responsive elements ( IREs ) in the 5′ UTR of ferritin mRNAs [22] . Expression of C . elegans ferritin genes is also sensitive to iron levels: iron supplementation increases ftn-1 expression , while iron chelation has the opposite effect . However , ftn-1 and ftn-2 lack IRE sequences in their 5′ UTRs and iron-dependent regulation seems to be achieved solely through transcriptional regulation [23] . The mechanism by which this occurs remains unknown , but iron-dependent regulation of ftn-1 requires a 63 bp iron-dependent element ( IDE ) in its promoter [20] . Research on the regulation of ftn-1 in C . elegans has contributed to our understanding of ‘restless leg syndrome’ , a human disease linked to iron deficiency in the brain . A haplotype of the gene MEIS1 has been associated with inheritance of the syndrome [24] but the gene's function was unknown . The involvement of the C . elegans ortholog unc-62 in regulating iron homeostasis was tested and a repressive role for this gene in ftn-1 regulation was identified . This regulation may be conserved in humans , since reduced MEIS1 expression seems to cause increased expression of human ferritin as well as of an iron transporter [25] . Thus , ftn-1 regulation in C . elegans can serve as a model for understanding the mechanisms of iron homeostasis in humans , and of human disease . In this study , we explore the biology of iron homeostasis in C . elegans by investigating further the regulation of ftn-1 . We show that ftn-1 is transcriptionally regulated by IIS/DAF-16 , and then perform a genetic screen using RNA mediated interference ( RNAi ) to identify factors influencing expression of a Pftn-1::gfp reporter . We identify several transcription factors known to act with IIS to regulate lifespan as factors that also regulate ftn-1 expression . We also reveal a major role for the hypoxia signaling pathway in ftn-1 regulation and iron homeostasis .
To ascertain whether ftn-1 expression might be regulated by insulin/IGF-1 signaling ( IIS ) and daf-16 , we examined published microarray-derived mRNA profiles comparing daf-2 and daf-16; daf-2 mutants [26] , [27] . These implied that ftn-1 mRNA levels are greatly elevated ( 47-fold increase ) in daf-2 compared to daf-16; daf-2 animals . This we were able to confirm using qRT-PCR ( Figure 1A ) . The increase in ftn-1 mRNA levels in daf-2 mutants was fully daf-16 dependent . Loss of daf-16 also decreased ftn-1 mRNA levels in daf-2 ( + ) animals . We then created a transgenic C . elegans line bearing a Pftn-1::gfp transcriptional reporter containing 3 . 8 kb of sequence upstream of the ftn-1 start codon . This was generated by microinjection of transgene DNA , and the resulting extrachromosomal transgene arrays were then chromosomally integrated . The Pftn-1::gfp transgene showed strong expression throughout the intestine , consistent with previous reports [20] . Effects of daf-2 and daf-16 upon Pftn-1::gfp expression paralleled those seen in ftn-1 mRNA levels ( Figure 1B , 1C ) . This confirms that ftn-1 is regulated by IIS , and shows that this regulation occurs principally in the intestine . We then used the Pftn-1::gfp reporter as the basis of an RNAi screen to investigate the mechanisms by which ftn-1 is regulated ( Figure 1D ) . The initial aim of this screen was to identify pathways that work coordinately with IIS , and regulatory factors that act downstream of DAF-16 . Expression of the integrated GFP ( green fluorescent protein ) reporter was intensified by mutation of daf-2 and sensitivity to RNAi was increased by introducing the rrf-3 ( pk1426 ) mutation . The resulting strain , of genotype rrf-3 ( pk1426 ) ; daf-2 ( m577ts ) ; wuIs177 [Pftn-1::gfp] , was raised at 15°C until the L4 stage , then transferred to RNAi plates and incubated at 25°C ( non-permissive temperature for daf-2 ( m577 ) ) . GFP fluorescence levels were measured in a plate-reader two days later . Given our interest in mechanisms of gene regulation , the RNAi screen was restricted to 812 genes encoding predicted transcription factors or other proteins associated with gene regulation [28] . RNAi of a number of these genes led to altered Pftn-1::gfp expression . In an initial screen , RNAi of 30 genes reduced Pftn-1::gfp expression by ≥20% ( Table S1 ) and we investigated these more thoroughly in several genetic backgrounds . For 10 of these genes , not including daf-16 , RNAi consistently and robustly reduced Pftn-1::gfp expression in multiple trials ( data not shown ) . We then verified the effect of RNAi on levels of mRNA from the endogenous ftn-1 gene . This identified four genes where RNAi robustly reduced ftn-1 mRNA levels: hsf-1 , mdl-1 , ada-2 and elt-2 ( Figure 2A , Table S1 ) . The heat-shock factor hsf-1 was previously shown to mediate effects of IIS on gene expression [29] . The screen also confirmed that the GATA transcription factor elt-2 plays a role in ftn-1 regulation . This is consistent with the role of elt-2 as an activator of intestinal gene expression [30]; moreover , elt-2 is the only previously described transcriptional activator of ftn-1 expression [20] . Thus , identification of hsf-1 and elt-2 in this unbiased screen is evidence of the efficacy of the screen . ada-2 encodes a homolog of the Ada2 subunit of various histone acetyl transferase ( HAT ) complexes that activate gene expression by modifying chromatin via histone acetylation [31] . It is possible that , ada-2 influences ftn-1 expression via effects on chromatin status . More notable is the identification of the MAD-like transcription factor mdl-1 as an activator of ftn-1 expression . mdl-1 plays a role in the protective effects of reduced IIS against a tumorous germline phenotype [32] and is upregulated in IIS mutants [10] , [26] , [32] . We confirmed that the null mutation mdl-1 ( tm311 ) reduces ftn-1 mRNA levels in daf-2 mutants ( Figure 2B ) . To explore whether these four factors might be acting downstream of DAF-16 , we tested whether RNAi reduces ftn-1 expression in a daf-16; daf-2 double mutant . The results imply that only MDL-1 does not require DAF-16 to activate ftn-1 expression . This suggests that mdl-1 acts downstream of daf-16 , or possibly in parallel to IIS , to regulate ftn-1 expression ( Figure 2C ) . Given that mdl-1 is a direct transcriptional target of DAF-16 [33] , the former seems more likely . Unexpectedly , RNAi of hsf-1 markedly increased ftn-1 expression in a daf-16; daf-2 background ( Figure 2C ) . The effects of hsf-1 RNAi ( Figure 2A ) imply that HSF-1 and DAF-16 act together to activate ftn-1 expression , as previously shown for other genes [29] . That loss of hsf-1 in daf-16; daf-2 mutants increases expression of ftn-1 could imply a repressive role of HSF-1 in the absence of DAF-16 . Alternatively , this increase might merely reflect a stressed state in the worms , caused by loss of both hsf-1 and daf-16 at 25°C ( see Discussion ) . Since ftn-1 is known to be responsive to iron levels , we also tested whether DAF-16 , HSF-1 or MDL-1 are required for iron-dependent regulation of ftn-1 . daf-16 , hsf-1 and mdl-1 mutants were treated with iron ( 25 mM ferric ammonium citrate , FAC ) and ftn-1 transcript levels measured by qRT-PCR . Iron-induced up-regulation of ftn-1 was unchanged in each case ( Figure S1A ) , i . e . these three factors do not mediate effects of iron on ftn-1 expression . RNAi of 28 genes further increased expression of the Pftn-1:gfp reporter ( Table S2 ) , already induced by daf-2 ( m577 ) . Of note was the large increase in expression upon RNAi of unc-62 , a transcription factor with a conserved role in ferritin regulation and , for its human ortholog , a possible role in the iron-related disorder ‘restless leg syndrome’ [25] . Also among the repressors of Pftn-1::gfp expression identified were hif-1 , encoding the hypoxia-inducible factor , and aha-1 , its binding partner ( HIFβ , also called ARNT ) . RNAi of either gene strongly increased Pftn-1::gfp expression , in the original daf-2 ( m577 ) ; Pftn-1::gfp strain ( Table S2 ) but also in two separate integrants of the Pftn-1::gfp reporter in a daf-2 ( + ) background ( Figure 3A , 3B and data not shown ) . We verified this activity of hif-1 by using the loss of function mutation hif-1 ( ia4 ) , which proved to greatly increase ftn-1 mRNA levels ( Figure 3C ) and Pftn-1::gfp expression ( Figure 3D ) . In a hif-1 ( ia4 ) mutant background , RNAi of aha-1 did not further increase Pftn-1::gfp expression ( Figure 3D ) , indicating that hif-1 and aha-1 act together to repress ftn-1 expression . The finding that hif-1 RNAi increases Pftn-1::gfp expression in a daf-2 mutant background suggests that hif-1 influences ftn-1 expression independently of IIS . Consistent with this , hif-1 or aha-1 RNAi increased Pftn-1::gfp expression in the absence of daf-16 ( Figure 3E ) . Results were similar at both 25°C and 20°C and at L4 and adult stages ( Figure 3E and data not shown ) . In addition , RNAi of hif-1 increased ftn-1 transcript levels in daf-16 mutants , and also in hsf-1 and mdl-1 mutants ( Figure S1B ) , indicating that none of these factors act downstream of HIF-1 to regulate ftn-1 expression . If HIF-1 is a repressor of ftn-1 expression , then elevation of HIF-1 levels should decrease expression of Pftn-1::gfp . Loss of vhl-1 ( von Hippel-Lindau factor ) leads to increased HIF-1 protein levels in C . elegans [34] . As expected , the deletion mutation vhl-1 ( ok161 ) markedly decreased expression of Pftn-1::gfp ( Figure 4A ) . Moreover , RNAi of vhl-1 reduced Pftn-1::gfp expression in hif-1 ( + ) but not hif-1 ( ia4 ) animals ( Figure 4B ) , and genetic deletion of vhl-1 led to a reduction in ftn-1 transcript levels that is also completely dependent on hif-1 ( Figure 4C ) . These results imply that HIF-1 acts downstream of VHL-1 as a repressor of ftn-1 expression . The prolyl hydroxylase EGL-9 hydroxylates the P621 residue of HIF-1 , which causes VHL-1 to bind to HIF-1 , leading to proteasomal degradation [34] . This hydroxylation reaction requires iron as a cofactor , suggesting that regulation of ftn-1 expression by iron might involve the HIF-1 pathway . We therefore tested whether the effects of iron on ftn-1 expression are hif-1 dependent . As expected given previous findings [23] , both expression of Pftn-1::gfp and ftn-1 mRNA levels were increased upon supplementation with iron ( ferric ammonium citrate , FAC ) and decreased upon treatment with the iron chelator 2′-2 bipyridil ( BP ) ( Figure 4D , 4E and Figure S2A and S2B ) . This is consistent with the previous observation that BP treatment greatly increases HIF-1 protein levels in C . elegans [34] , since increased HIF levels would be expected to further repress ftn-1 expression . By contrast , in hif-1 ( ia4 ) mutants , addition of iron did not increase either Pftn-1::gfp expression or ftn-1 mRNA levels . This implies that hif-1 mediates the induction of ftn-1 expression by iron . Iron chelation did not decrease ftn-1::gfp and ftn-1 expression in hif-1 ( ia4 ) , but instead increased it . The cause of this induction remains unexplained . One possibility is that BP treatment leads to cellular stress and induction of other stress response regulators ( e . g . DAF-16 ) , which can activate ftn-1 expression in the absence of the repressive effects of HIF-1 ( see Discussion ) . In order to investigate whether HIF-1 represses ftn-1 expression by directly binding to the ftn-1 promoter , we carried out a chromatin immunoprecipitation ( ChIP ) assay using C . elegans expressing Myc-tagged HIF-1 [35] and an anti-Myc antibody . We used three lines: wild type ( N2 ) , ZG429 [hif-1::myc] and GA654 [hif-1::myc vhl-1 ( ok161 ) ] . Given that vhl-1 mutants have elevated HIF-1 levels and reduced ftn-1 mRNA levels ( Figure 4C ) , greater levels of HIF-1::Myc binding to the ftn-1 promoter should be detectable in vhl-1 mutants , if the interaction is in fact direct . We first checked that our ChIP protocol allowed us to measure binding by HIF-1::Myc by testing binding to the promoter of a known HIF-1 target gene , nhr-57 . We designed one set of primers to amplify the region of the promoter containing two putative hypoxia response elements ( HREs ) and another set of primers targeting an area within the 3′ UTR of this gene . Quantity of qRT-PCR amplified promoter DNA was then compared to the 3′ UTR quantity as a test of enrichment of the promoter in our ChIP DNA pools . This amplification from the 3′ UTR ( to which HIF-1 is not expected to bind ) allowed us to control for input quantity . We saw a large ( 8 . 3-fold ) enrichment of the nhr-57 promoter sequence in the HIF-1::Myc lines and an even greater ( 14 . 9-fold ) enrichment when HIF-1::Myc was stabilized by deletion of vhl-1 ( Figure 4F ) . Relative amounts of HIF-1::Myc were monitored by Western blotting of the same ChIP samples using the same aliquot of anti-Myc antibody used for ChIP , and we were able to confirm that vhl-1 ( ok161 ) increases HIF-1::Myc protein levels ( Figure 4F ) . We then measured binding to the ftn-1 promoter through qRT-PCR against the promoter sequence of ftn-1 . For this , we used a primer pair specific to the IDE sequence . These results were again normalized against the same nhr-57 3′UTR in order to correct for differences in input quantity . While weaker than binding to Pnhr-57 , enrichment of the Pftn-1 sequence in HIF-1::Myc and stabilized HIF-1::Myc lines was statistically significantly different to that seen in wild-type controls ( Figure 4F ) . This is evidence that HIF-1 represses ftn-1 expression through direct binding to its promoter . The repression of ftn-1 expression by HIF-1 and the requirement for iron in the degradation of HIF-1 by the proteasome suggests a possible mechanism for the iron-dependent regulation of ftn-1 in which changes in iron levels alter the level of HIF-1 protein , which in turn alter ftn-1 expression . Since the iron-dependent degradation of HIF-1 occurs via the action of VHL-1 , HIF-1 protein levels in C . elegans are not sensitive to iron levels when VHL-1 is absent [36] . We found that loss of vhl-1 largely abrogated the induction of Pftn-1::gfp expression by iron supplementation , though there was still a significant induction of lesser magnitude ( Figure 5A ) . Reduction of Pftn-1::gfp expression by iron chelation was not affected by loss of vhl-1 ( Figure 5B ) . Taken together , this suggests that regulation of ftn-1 by iron may occur partially , but not exclusively , through changes in HIF-1 protein levels regulated by iron-dependent degradation . The induction of ftn-1 levels by iron requires a 63 bp element ( the iron-dependent element , or IDE ) in the gene's promoter [20] . We wondered whether the hif-1 pathway might mediate the effects of iron on IDE-mediated gene expression . A reporter strain carrying a ftn-1 promoter lacking the IDE is insensitive to changes in iron levels [20] . Using these same reporters we found that absence of the IDE abolished hif-1 RNAi-induced induction of expression ( Figure 5C ) . Another reporter construct with just the IDE sequence fused to a minimal promoter and driving GFP expression was previously shown to be responsive to iron [20] . We found that loss of hif-1 increased ide::gfp expression , demonstrating that hif-1 does promote gene expression from the IDE ( Figure 5D ) . Moreover , addition of iron did not induce ide::gfp expression in hif-1 mutants ( Figure 5D ) . However , in hif-1 mutants treatment with the iron chelator BP still reduced ide::gfp expression ( Figure 5E ) . This possibly reflects an effect of BP on ftn-1 that is independent of its effects on iron levels , or the existence of a second iron-dependent factor . These results show that the IDE is subject to regulation by HIF-1 and suggest that HIF-1 mediates the effects of iron on IDE-mediated gene expression . As previously described , loss of vhl-1 decreases expression of Pftn-1::gfp ( Figure 4A ) . This is expected given that HIF-1 represses ftn-1 expression and that loss of vhl-1 increases HIF-1 levels [34] . The prolyl hydroxylase EGL-9 targets HIF-1 for proteasomal degradation , and loss of egl-9 causes a similarly large increase in HIF-1 protein levels as loss of vhl-1 [36] . We therefore expected that loss of egl-9 , like that of vhl-1 , would reduce Pftn-1::gfp expression . In fact , deletion of egl-9 caused an 11-fold increase in Pftn-1::gfp expression ( Figure 6A ) and a ∼950-fold increase in ftn-1 mRNA levels ( Figure 6B ) . Animals with a different allele , egl-9 ( n586 ) , also showed increased ftn-1 mRNA levels ( Figure S3A ) . Visible Pftn-1::gfp expression remained restricted to the intestine in wild type , vhl-1 mutants and egl-9 mutants . vhl-1-independent effects of EGL-9 on HIF-1 target gene expression have been observed previously [36] . Our findings suggest that in the case of ftn-1 regulation , egl-9 can act independently of , and antagonistically to , vhl-1 . As expected , loss of egl-9 induced ftn-1 expression even in the absence of vhl-1 ( Figure S3B and S3C ) . However , egl-9 RNAi did not increase ftn-1 transcript or Pftn1::gfp expression in the absence of hif-1 ( Figure 6B and Figure S3D ) . This implies that the inhibition of ftn-1 expression by EGL-9 also requires hif-1 . Thus , egl-9 and vhl-1 inhibit and activate expression of ftn-1 , respectively , and both activities require hif-1 . One possibility is that EGL-9 inhibits ftn-1 expression by stimulating HIF-1 activity via an as yet unidentified pathway .
IIS and DAF-16 play a pivotal role in the organismal decision between growth and diapause . Under growth-promoting conditions , DAF-16 is inactivated through cytoplasmic retention , which facilitates reproductive growth [8] , [37] . In the absence of sufficient food or given exposure to certain forms of stress , DAF-16 enters the nucleus and transcriptionally specifies a survival program . This entails delayed reproduction , enhanced stress resistance and increased lifespan . Modulation of DAF-16 activity is therefore crucial for ensuring an optimal response to the worm's environment; with growth and reproduction under conditions that are propitious to growth , and developmental arrest , stress protection and increased longevity under conditions that are not . Regulation of ftn-1 by DAF-16 suggests the existence of a trade-off between growth and stress resistance involving iron homeostasis . A role for ferritin in regulating growth via its effects on iron homeostasis has been described previously in mammalian cells [38] . This study found that Myc , a bHLH transcription factor with a major role in promoting cellular proliferation , can repress H-ferritin expression . Overexpression of ferritin in cells carrying activated Myc led to a decrease in in vitro clonogenicity , and this effect could be rescued by addition of iron , suggesting that Myc–mediated repression of ferritin expression favors growth by increasing iron availability . The study identified DNA synthesis as a possible mechanism for iron-dependent control of cellular proliferation by c-Myc , as DNA synthesis is increased by c-Myc in a manner dependent on ferritin repression and the associated increases in iron availability . This finding is consistent with the requirement for iron in the activity of ribonucleotide reductase , the rate-limiting enzyme in DNA synthesis . Similar mechanisms may be at play in the regulation of ferritin expression by IIS . When conditions favor growth , and IIS is increased , reduced ftn-1 expression is expected to increase iron availability , thus fulfilling a key requirement for growth . While free iron is required for growth , it can also cause harm by catalyzing the Fenton reaction , which increases levels of ROS and molecular damage . When conditions are not suitable for growth , IIS is reduced , and increased ftn-1 expression is expected to lower levels of free iron and of ROS , thereby protecting against stress . Consistent with this , induced over-expression of ftn-1 causes resistance to oxidative stress ( S . Valentini and D . Gems , unpublished results ) . Thus , upregulation of ftn-1 likely contributes to the broader increase in cytoprotection seen when IIS is reduced . Reduced IIS also increases levels of autophagy in C . elegans [39] , [40] and autophagy releases iron from ferruginous materials , such as mitochondrial metalloproteins [41] . This predicts that reduced IIS will increase free iron levels , and concomitant elevation of ftn-1 expression could ensure that iron released by autophagy does not cause molecular damage . Using an RNAi screen we identified new regulators of ftn-1 , including hsf-1 and mdl-1 . It was previously shown that in daf-2 mutants the heat shock factor HSF-1 acts in concert with DAF-16 to promote expression of small heat shock proteins and other molecular chaperones , which contribute to longevity [29] . We find that hsf-1 is also involved in the induction of ftn-1 in daf-2 mutants , since loss of hsf-1 reduced ftn-1 expression in daf-2 but not daf-16; daf-2 mutants . The MAD-like transcription factor mdl-1 is also regulated by IIS . Microarray and qRT-PCR studies showed it to be up-regulated in daf-2 mutants [10] , [26] , [32] . mdl-1 also contributes to the resistance of daf-2 mutants to germline tumor formation in the gld-1 tumor model , and to daf-2 mutant longevity [32] . That MDL-1 activates ftn-1 expression is consistent with the role of mammalian MAD as an inhibitor of Myc , which represses ferritin expression ( see above ) ; however , C . elegans does not possess any clear ortholog of Myc [42] , [43] . A study of DAF-16 binding sites did not provide evidence that ftn-1 is a direct regulatory target of DAF-16 [33] , but suggested that mdl-1 might be . Given that ftn-1 may be a direct target of MDL-1 [44] , [45] , one possibility is that activation of mdl-1 expression by DAF-16 leads to increased ftn-1 expression . This hypothesis predicts that abrogation of mdl-1 expression should decrease ftn-1 expression more in daf-2 than in daf-16; daf-2 animals , but this is not the case ( Figure 2A , 2C ) . This could imply that mdl-1 regulates ftn-1 independently of daf-16 , at least in part . We discovered that loss of hif-1 or its binding partner aha-1 greatly increased ftn-1 expression in daf-2 mutants . This implicated hypoxia signaling in the regulation of ftn-1 . The HIF transcription factor is composed of an α and a β subunit , encoded by the genes hif-1 and aha-1 in C . elegans . HIF regulates the transcriptional response to hypoxia in both worms and vertebrates and , as expected , worms lacking hif-1 are hypersensitive to hypoxia [46] . Levels of HIFβ protein are relatively stable , whereas HIFα is constantly being degraded by the proteasome under normal , non-hypoxic conditions . In both worms and higher organisms , this occurs because the HIFα/HIF-1 protein is hydroxylated at conserved proline residues by prolyl hydroxylase ( PHD ) , encoded by the egl-9 gene in worms . After hydroxylation by PHD/EGL-9 , the von Hippel-Lindau protein VHL-1 binds to HIFα , which targets it for degradation [34] , [47] . PHDs require oxygen , iron and 2-oxoglutarate for the hydroxylation reaction . When cells are kept under hypoxic conditions or when an iron chelator is added , the proline residue in HIFα is not hydroxylated and the HIFα protein accumulates [48] . That loss of hif-1 has such dramatic effects on gene expression under normoxic conditions demonstrates that HIF-1 affects gene regulation even at the very low levels of HIF-1 found when it is being hydroxylated and degraded . Similarly , it was previously observed that loss of hif-1 can increase C . elegans lifespan under normoxic conditions [49] . Consistent with this , we find statistically significant levels of binding of the non-stabilized HIF-1::Myc protein to both ftn-1 and nhr-57 promoters ( Figure 4F ) . Since iron is a required cofactor for hydroxylation of HIF by PHD , levels of iron affect those of HIF . For example , in C . elegans , depletion of iron using the iron chelator 2-2′ bipyridyl stabilizes HIF-1 [34] , and feeding mice a low-iron diet leads to increased HIFα levels [50] . This increase in HIF-1 levels is not without consequence: chelation of iron has also been shown to increase expression of the C . elegans HIF-1 target gene nhr-57 , indicating that the stabilization of HIF upon loss of iron leads to HIF-1-dependent changes in gene expression [51] . In vertebrates , HIF activates expression of genes involved in regulating iron homeostasis , including heme oxygenase [52] , the transferrin receptor [53] , [54] , ceruloplasmin [55] , DMT1 [56] and possibly ferroportin [57] . Loss of HIF-2α in mice causes decreased iron levels in the plasma and livers of mice [56] . It has therefore been suggested that HIF can act as an iron sensor: low iron levels lead to HIF stabilization , which leads to changes in gene expression that increase iron levels [57] . The results presented here support this hypothesis . The repression of ferritin expression by hif-1/aha-1 is consistent with a role of HIF in increasing iron availability . By this view , lower ferritin expression upon HIF activation would reduce iron storage capacity , thereby increasing iron availability . We therefore investigated whether HIF mediates iron-dependent regulation of ftn-1 , and this proved to be the case: ftn-1 regulation by iron is blocked in hif-1 mutants . In wild-type animals iron supplementation increases ftn-1 expression while iron depletion decreases it . By contrast , in hif-1 mutants iron supplementation does not increase ftn-1 expression . Treatment of hif-1 mutants with the iron chelator 2-2′ bipyridyl ( BP ) caused a large increase , rather than decrease , of ftn-1 expression . This was unexpected , but we noticed that BP treated worms were somewhat sickly in appearance . One possibility is that toxicity of BP in hif-1 mutants triggers other stress response mediators ( e . g . DAF-16 ) that activate ftn-1 expression . This is consistent with our observation that stressful conditions tend to induce expression of this reporter . Similar to treatment with BP , RNAi of hsf-1 in daf-16 ( mgDf50 ) ; daf-2 ( m577 ) animals raised at 25°C also caused the worms to have a sickly appearance and induced Pftn-1::gfp expression ( Figure 2C ) . Moreover , we observed that starved animals also show elevated Pftn-1::gfp expression ( data not shown ) . The requirement for hif-1 in the iron-dependent regulation of ftn-1 suggests that this regulation may occur via iron-dependent degradation of HIF-1 . However , our data implies that this is not the whole story . Mutants of vhl-1 have constitutively stabilized HIF-1 and its levels cannot therefore respond to changes in iron ( or oxygen ) levels [34] , [36] . While the increase in Pftn-1::gfp expression upon treatment with iron was greatly reduced in vhl-1 mutants , Pftn-1::gfp expression was still elevated compared to the control treatment . This implies that iron-dependent degradation of HIF-1 is not the sole mechanism by which ftn-1 is regulated in response to iron levels . The control of ftn-1 expression by iron was previously shown to be mediated by the 63 bp iron-dependent element ( IDE ) in the ftn-1 gene promoter [20] . This implied the presence of an unknown iron-responsive transcriptional activator exerting effects upon the IDE . Our findings strongly suggest that this factor is HIF . We found that loss of hif-1 increases ide::gfp expression . Moreover , in the absence of hif-1 , iron supplementation failed to induce ide::gfp expression . Furthermore , Romney et al . ( 2008 ) identified three conserved elements ( called DR elements ) , with the consensus sequence: CACGTA ( C/G ) ( C/A/G ) in the IDE to which they attribute the responsiveness of ftn-1 expression to iron levels . This DR sequence has homology to the E-box motif , which led Romney et al . to suggest that the iron-sensory pathway includes a basic helix-loop-helix ( bHLH ) transcription factor . Both HIF-1 and AHA-1 belong to this family of proteins . In fact , the conserved DR sequence described by Romney et al . contains the putative C . elegans hypoxia response element ( HRE ) [58] ( in reverse orientation ) . Moreover , using ChIP , we found that epitope-tagged HIF-1 bound to the region of the promoter containing the IDE . Taken together , these results support the view that HIF-1 acts as an iron sensor in C . elegans , suppressing ftn-1 expression by binding to the IDE , although the mechanism by which iron levels are detected has not yet been identified . Our discovery of the role of HIF in iron homeostasis in C . elegans has notable implications in terms of the evolution of HIF as an iron sensor . The effects of iron on HIF levels in higher organisms has been viewed in the context of HIF's role in stimulating erythropoiesis . Since erythropoiesis requires large quantities of iron , it was proposed that the purpose of the HIF-mediated induction of genes involved in increasing iron availability is to supply iron for erythropoiesis [57] . That HIF regulates iron homeostasis in nematodes implies that the evolution of this function predates the emergence of a circulatory system . The sensitivity of hypoxia signaling to oxygen , iron and ROS , which interact and produce oxidative damage to the cell , further suggests that HIF may have an ancestral role in fine-tuning the response to different levels of these potentially toxic substances . Against expectation , loss of egl-9 increased expression of Pftn-1::gfp , rather than decreasing it . This does not merely indicate that EGL-9 regulates targets other than HIF-1 , since the induction is hif-1 dependent . Given that loss of egl-9 or vhl-1 cause similar increases in HIF-1 protein levels [36] , this finding suggests that increased HIF protein levels can be associated with both increased and decreased ftn-1 expression . That both decreased expression upon vhl-1 deletion and increased expression upon egl-9 deletion require hif-1 is difficult to reconcile . However , vhl-1 independent effects of EGL-9 on HIF-1 target gene expression have been observed previously [36] . HIF-1 target genes are often more highly induced by loss of egl-9 than of vhl-1 , despite identical levels of HIF-1 stabilization in each case [36] . This implies that regulation of HIF-1 target gene expression by EGL-9 occurs via both VHL-1-dependent and independent mechanisms . Additionally , mutations in egl-9 protect worms against infection by Pseudomonas aeruginosa and this effect is dependent on HIF-1 . However , stabilization of HIF-1 by other means is insufficient to achieve this effect , again showing that EGL-9 can act via mechanisms other than HIF-1 stabilization [59] . We find that the effects of loss of egl-9 on ftn-1 do not require vhl-1 either . But in contrast to the other examples cited above , regulation of ftn-1 by the vhl-1-dependent and independent pathways downstream of EGL-9 act antagonistically , the former repressing ftn-1 expression and the latter activating it . Thus , our findings imply that EGL-9 not only represses HIF-1 activity by the well-characterized VHL-1-dependent pathway , but also modulates HIF-1 activity by an unknown mechanism ( Figure 7 ) . Prolyl hydroxylases are sensitive proteins capable of responding not only to iron and oxygen levels , but also to cues from metabolism [60] and to ROS [61] , [62] . One or more of these may trigger the VHL-independent activity of EGL-9 . Previous studies have suggested that hypoxia and IIS might act in concert to regulate gene expression . daf-2 mutants are highly resistant to hypoxia [63] and microarray studies found an over-representation of genes containing hypoxia response elements ( HRE ) among IIS/DAF-16 regulated genes [26] . A study using murine embryonic fibroblasts found that the FOXO3a transcription factor inhibits HIF-1 mediated gene regulation [64] . We were therefore interested in investigating whether hypoxia signaling and IIS interact to regulate ftn-1 . One model for the joint regulation of ftn-1 by hif-1 and IIS/DAF-16 that we initially considered is that loss of hif-1 activates DAF-16 which in turn activates ftn-1 . DAF-16 is a stress-responsive transcription factor , so it seemed possible that stress caused by loss of HIF triggers a DAF-16-mediated cytoprotective response . There is evidence that loss of hif-1 does indeed have this effect on DAF-16: hif-1 mutants are long-lived and this lifespan extension has been shown to require daf-16 [49] . However , several observations argue against the idea that DAF-16 mediates HIF-1 effects . Firstly , HIF-1 binds directly to the ftn-1 promoter ( Figure 4F ) . Secondly , the effects of loss of hif-1 on ftn-1 expression do not require daf-16 ( Figure 3E ) . Finally , loss of hif-1 can further induce the expression of Pftn-1::gfp in daf-2 mutants . The fact that daf-16 and hif-1 have opposite effects on ftn-1 expression bears consideration . Recent reports have found that HIF-1 overexpression extends lifespan [49] , suggesting that HIF-1 activity has a similar effect to increased DAF-16 activity . Whether this occurs through the activation of a similar set of genes is unknown . Our finding that HIF-1 and DAF-16 can have opposing effects on gene expression suggests that the relationship between the two gene-sets is complex . Regulation of ftn-1 by HIF-1 and DAF-16 could be a special case in which DAF-16 regulation occurs as part of a broad response to lower oxidative stress whereas HIF-1 acts as an iron sensor . Further work is required in order to establish whether antagonistic regulation by DAF-16 and HIF-1 is specific to ftn-1 or whether it represents a more general pattern of gene regulation by the two pathways . In summary , this study maps out a complex gene-regulatory network controlling expression of ftn-1 and , by extension , iron homeostasis in the nematode C . elegans . This reveals the acute sensitivity of iron homeostasis to environmental conditions , allowing fine tuning of iron availability in the face of variability of factors that increase free iron ( increased environmental iron , growth arrest , increased autophagy ) and decrease it ( reduced environmental iron , increased growth ) . Our results also underscore the value of C . elegans as a model system for understanding mammalian iron homeostasis , and the pathologies that can result from its breakdown .
Maintenance and culture of C . elegans was carried out as published [65] , [66] , [67] . The following strains were used: CB5602 vhl-1 ( ok161 ) , DR1563 daf-2 ( e1370 ) , DR1567 daf-2 ( m577 ) , GA300 daf-16 ( mgDf50 ) ; daf-2 ( m577 ) , GA633 daf-2 ( m577 ) ; wuIs177 [Pftn-1::gfp lin-15 ( + ) ] , GA636 rrf-3 ( pk1426 ) ; daf-2 ( m577 ) ; wuIs177 [Pftn-1::gfp lin-15 ( + ) ] , GA639 daf-16 ( mgDf50 ) ; wuIs177 [Pftn-1::gfp lin-15 ( + ) ] , GA640 wuIs176 [Pftn-1::gfp lin-15 ( + ) ] , GA641 wuIs177 [Pftn-1::gfp lin-15 ( + ) ] , GA642 hif-1 ( ia4 ) ; wuIs177 [Pftn-1::gfp lin-15 ( + ) ] , GA643 daf-16 ( mgDf50 ) ; daf-2 ( m577 ) ; wuIs177 [Pftn-1::gfp lin-15 ( + ) ] , GA654 unc-119 ( ed3 ) vhl-1 ( ok161 ) iaIs128[Phif-1::hif-1a::myc unc-119 ( + ) ] , GA675 xtEx79 [Δpes-10 ( +63 ) ::GFP-his , pha-1 ( + ) ] , GA676 hif-1 ( ia4 ) xtEx79 [Δpes-10 ( +63 ) ::GFP-his , pha-1 ( + ) ] , GA688 pha-1 ( e2123ts ) xtEx79 [Δpes-10 ( +63 ) ::GFP-his , pha-1 ( + ) ] , GA688 pha-1 ( e2123ts ) ; hif-1 ( ia4 ) xtEx79 [Δpes-10 ( +63 ) ::GFP-his , pha-1 ( + ) ] , GA694 wuIs176 [Pftn-1::gfp lin-15 ( + ) ] egl-9 ( sa307 ) ] , GA1200 mdl-1 ( tm311 ) , GA1203 daf-2 ( e1370 ) ; mdl-1 ( tm311 ) , GA1204 daf-2 ( m577 ) ; mdl-1 ( tm311 ) , GR1307 daf-16 ( mgDf50 ) , JT307 egl-9 ( sa307 ) , N2 , PS3551 hsf-1 ( sy441 ) , UZ96 pha-1 ( e2123ts ) xtEx79 [Δpes-10 ( +63 ) ::GFP-his , pha-1 ( + ) ] , XA6900 pha-1 ( e2123ts ) qaEx6902 [Pftn-1 ( Δ63 ) ::[Δpes-10::GFP-his , pha-1 ( + ) ] , XA6902 pha-1 ( e2123ts ) qaEx6902 [Pftn-1::[Δpes-10::GFP-his , pha-1 ( + ) ] and ZG31 hif-1 ( ia4 ) . ZG429 unc-119 ( ed3 ) iaIs128[Phif-1::hif-1a::myc unc-119 ( + ) ] Worms were maintained at 20°C unless otherwise indicated . Multiple mutants were created using standard methodologies and the presence of genomic deletions was tested via PCR . Genotyping was carried out by lysis of parent animals using proteinase K ( Sigma ) and subsequent PCR using the following primers . For daf-16 ( mgDf50 ) : daf-16F1 , gccactttattggaatttgagc; and daf-16R1 , atcctcccatagaaggaccatt . For hif-1 ( ia4 ) : hif-1_ex_fwd1 , gctcctcctactccacctttg , hif-1_ex_rev1 , gtgacgagttgtgaatgcacc , hif-1_int_rev1 . 2 , tcggcgatggtgtcttcagtc . For rrf-3 ( pk1426 ) : rrf-3_ex_fwd1 , gagttcgcatcaagtttcac , rrf-3_ex_rev1 , tgccttcgtacatttcaacc and rrf-3_int_rev2 , ggtatttattgcttcctgccac . For vhl-1 ( ok161 ) : DA75 , gctgtcaatcggagcactgtc , DA76 , ttgctgaggtctctggggtc , and DA77 , gttagctctgccacgaatacgatg . For egl-9 ( sa307 ) : DA117 , acaaagacaggtgttgcgaatgag , DA118 , ttgtagtgatccgagcccag , and DA119 , gatgcttctgatgttcttggagg . The promoter::gfp transgene of ftn-1 was created using methods as previously described [68] and the transgenic strain was created by microinjection . The primers used for creation of the construct were: ftn-1 . 5'ex , tgcttactggttctgccgag , ftn-1 . 5'in , tgtagggtttgattgtggtttg , ftn-1 . 3'fus , agtcgacctgcaggcatgcaagctttgacgagctagagacatgac . Extrachromosomal arrays were integrated by X-ray irradiation . The method used to quantify GFP expression was adapted from one used in an earlier study [69] . Using a worm pick , samples of forty adult worms were transferred into the wells ( V-shaped ) of microtitre plates ( Greiner ) . Fluorescence was then measured in a GeniosPlus plate reader ( Tecan ) at wavelengths appropriate for GFP ( excitation: 495 nm; emission: 535 nm ) using a fixed gain of 75 . Quantification of GFP expression from transgenes with low level expression was carried out using a Leica DMRXA2 microscope using a GFP filter cube ( excitation: 470/40 nm; emission: 525/50 nm ) , an Orca C10600 digital camera ( Hamamatsu ) and Volocity image analysis software ( Improvision ) . The transcription factor RNAi library used for this project was generously provided by Dr . Weiqing Li ( University of Washington ) . Similar libraries are now available commercially ( geneservice . co . uk ) . Where RNAi robustly affected ftn-1 expression levels , RNAi plasmid inserts were sequenced to confirm their identity using the primers JJM130 ( gggaagggcgatcggtgcgggcc ) and JJM131 ( gcgcagcgagtcagtgagcgagg ) . RNA was isolated from 2-day old adults after three washes , which removed E . coli and L1 progeny from the sample . After RNA isolation cDNA was synthesized using SuperScript II reverse transcriptase ( Invitrogen ) using oligo dT ( Invitrogen ) . qRT-PCR was carried out using Fast SYBR Green Master Mix ( Applied Biosystems ) and the 7900 HT Fast PCR system ( Applied Biosystems ) . Normalization of transcript quantity was carried out using the geometric mean of three stably expressed reference genes Y45F10D . 4 , pmp-3 , and cdc-42 in order to control for cDNA input , as previously described [70] . The following primers were used for this assay . Y45F10D . 4: DA90 , gtcgcttcaaatcagttcagc , and DA91 , gttcttgtcaagtgatccgaca . pmp-3: DA88 , gttcccgtgttcatcactcat , and DA89 , acaccgtcgagaagctgtaga . cdc-42: DA86 , ctgctggacaggaagattacg , and DA87: ctcggacattctcgaatgaag . ftn-1: ftn-1_fwd_RT2 , cggccgtcaataaacagattaacg , and ftn-1_rev_RT2 cacgctcctcatccgattgc . qRT-PCR of ChIP DNA pools was carried out for the nhr-57 promoter using DA130: cctcccgcgtctccacattcaatc and DA131: cagcgaggtctgggttttccg , the nhr-57 3′UTR using DA135: tggcacaagatatgacgaaagctg and DA136: ggcgagaaatttgttgtaggttgcc , and the ftn-1 promoter using DA139: aacagctcacgtagccaatgataag and DA140: gcatcacatgagctgcccta . All results shown are the mean of at least three independent biological replicates and error bars represent the s . e . m . Statistical significance was calculated by two-way or one-way ANOVA of either raw values or log-transformed quantities , depending on circumstances . The protocol for chromatin immunoprecipitation was adapted from Mukhopdhyay et al . [71] . C . elegans cultures were grown for two generations in S-media with suspended OP50 at 20°C with constant shaking at 200 rpm . The worms were collected and washed four times in PBS buffer and then re-suspended in PBS containing 1% formaldehyde . Samples were then partially lysed using 8 strokes with a 1/3 turn in a 7 cm Dunce homogenizer and then incubated for 17 minutes with gentle mixing at room temperature . Crosslinking was stopped by addition of 200 µl 2 . 5 mol/L Glycine solution and 20 minutes further incubation at room temperature . After four washes in PBS containing protease inhibitor tablets ( Complete , Roche ) , samples were flash frozen and stored at −80°C . After thawing , 2 mL of HLB buffer [50 mM HEPES-KOH , pH 7 . 5 , 150 mM NaCl , 1 mM EDTA , 0 . 1% ( wt/vol ) sodium deoxycholate , 1% ( vol/vol ) Triton X-100 , 0 . 1% ( wt/vol ) SDS and 1× Complete protease inhibitor] was added and sonication was carried out at 70% intensity for 7 bursts of 30 seconds in the Vibracell sonicator ( Sonics ) . Protein quantity was estimated by Bradford assay ( Biorad ) and 2 mg were diluted into to 500 µl of in HLB buffer . Three 50 µl aliquots were removed at this point . DNA isolated from these samples was subsequently used as input controls . Samples were precleared for 1 h in prewashed salmon sperm DNA/protein-A agarose beads ( Millipore ) and then incubated overnight with 10 µl of anti-Myc Ab ( 9b11; Cell signalling ) . Samples were then incubated with prewashed salmon sperm DNA/protein-A agarose beads for 2 h . The beads were then washed twice in WB1 [50 mM HEPES-KOH , pH 7 . 5 , 150 mM NaCl , 1 mM EDTA , 1% ( wt/vol ) sodium deoxycholate , 1% ( vol/vol ) Triton X-100 , 0 . 1% ( wt/vol ) SDS and 1× Complete protease inhibitor] , twice in WB2 [50 mM HEPES-KOH , pH 7 . 5 , 1 M NaCl , 1 mM EDTA , 1% ( wt/vol ) sodium deoxycholate , 1% ( vol/vol ) Triton X-100 , 0 . 1% ( wt/vol ) SDS and 1× Complete protease inhibitor] and once in WB3 [50 mM Tris-HCl , pH 8 , 0 . 25 mM LiCl , 1 mM EDTA , 0 . 5% ( vol/vol ) NP-40 and 0 . 5% ( wt/vol ) sodium deoxycholate] . Crosslinking was reversed by addition of proteinase K solution [50 mM Tris-HCl , pH 8 , 25 mM EDTA , 1 . 25% ( wt/vol ) SDS , 160 µg/ml proteinase K ( Qiagen ) ] and incubation for 2 h at 45°C and overnight at 65°C . DNA was isolated by applying solution to Qiagen PCR purification columns ( Qiagen ) . | Iron plays a role in many biological processes , including energy generation and DNA replication . But to maintain health , levels of cellular iron must be just right: too much or too little iron can cause illnesses , such as anemia and hemochromatosis , respectively . Animals therefore carefully control their iron levels by regulating of iron uptake , transport , and storage within protein capsules called ferritins . But how do they coordinate this ? Using the model organism C . elegans , we have discovered a network of genes and pathways that control iron homeostasis . We find that ferritin is regulated by insulin/IGF-1 signaling , which also controls growth and resistance to oxidative stress in response to harsh environmental conditions . Ferritin is also regulated by the hypoxia signaling pathway , which responds to oxygen and iron levels as well as to metabolic cues . We find that the hypoxia pathway acts as an iron sensor , a role it may also play in humans . Our work defines a network of signaling pathways that can adjust iron availability in response to a range of environmental cues . Understanding this network in C . elegans can help us to understand the causes of iron dyshomeostasis in humans , which can profoundly affect health . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
]
| [
"biochemistry",
"genetics",
"biology",
"metabolism",
"genetics",
"and",
"genomics"
]
| 2012 | Insulin/IGF-1 and Hypoxia Signaling Act in Concert to Regulate Iron Homeostasis in Caenorhabditis elegans |
Processing of unattended threat-related stimuli , such as fearful faces , has been previously examined using group functional magnetic resonance ( fMRI ) approaches . However , the identification of features of brain activity containing sufficient information to decode , or “brain-read” , unattended ( implicit ) fear perception remains an active research goal . Here we test the hypothesis that patterns of large-scale functional connectivity ( FC ) decode the emotional expression of implicitly perceived faces within single individuals using training data from separate subjects . fMRI and a blocked design were used to acquire BOLD signals during implicit ( task-unrelated ) presentation of fearful and neutral faces . A pattern classifier ( linear kernel Support Vector Machine , or SVM ) with linear filter feature selection used pair-wise FC as features to predict the emotional expression of implicitly presented faces . We plotted classification accuracy vs . number of top N selected features and observed that significantly higher than chance accuracies ( between 90–100% ) were achieved with 15–40 features . During fearful face presentation , the most informative and positively modulated FC was between angular gyrus and hippocampus , while the greatest overall contributing region was the thalamus , with positively modulated connections to bilateral middle temporal gyrus and insula . Other FCs that predicted fear included superior-occipital and parietal regions , cerebellum and prefrontal cortex . By comparison , patterns of spatial activity ( as opposed to interactivity ) were relatively uninformative in decoding implicit fear . These findings indicate that whole-brain patterns of interactivity are a sensitive and informative signature of unattended fearful emotion processing . At the same time , we demonstrate and propose a sensitive and exploratory approach for the identification of large-scale , condition-dependent FC . In contrast to model-based , group approaches , the current approach does not discount the multivariate , joint responses of multiple functional connections and is not hampered by signal loss and the need for multiple comparisons correction .
Faces with a fearful expression are thought to signal the presence of a significant , yet undetermined source of danger within the environment , or ‘ambiguous threat’ [1] . Evidence from fMRI and evoked potentials ( ERPs ) suggest that fearful face processing can strongly affect brain systems responsible for face recognition and memory during implicit ( consciously perceived but unattended ) presentation of these stimuli [2] , [3] . Group-based fMRI studies have shown that the perception and processing of facial emotional expression engages multiple brain regions including the fusiform gyrus , superior temporal sulcus , thalamus , as well as affect-processing regions such as amygdala , insula , anterior cingulate cortex among others [4]–[7] . However , to the authors' knowledge , no study to date has identified features of brain activity that contain sufficient information to reliably decode , or “brain-read” , the threat-related emotional expression of unattended ( implicitly perceived ) faces within individual subjects . The identification of such features , though less well quantified as in group model-based studies , would have a greater capacity for representing distinctions between different cognitive-emotional perceptual states [8] , and hence could contribute in advancing our understanding of the neural mechanisms that underlie threat detection and facial emotion processing . Most group fMRI approaches that have studied the neural correlates of emotional face perception have relied on univariate approaches [9]–[11] which identify regions correlated with a regressor-of-interest , but ignores any interactions with other regions . Bivariate approaches have been applied , but assess the interactivity ( functional connectivity ) of only one seed region ( usually amygdala ) with the rest of the brain [12] , [13] . Even though several notable studies have taken a multivariate approach in assessing the effective connectivity among multiple brain regions during emotional face processing [14]–[16] , a limited number of nodes were included in the networks and they were selected based on a priori anatomical knowledge or on their activation in conventional , General Linear Model ( GLM ) -based mass univariate analyses . However , univariate GLM approaches make strong assumptions about the hemodynamic response ( i . e . sustained periods of activation or deactivation relative to baseline ) , while functional connectivity offers a complementary and more data-driven and exploratory measure that makes use of temporal correlations to estimate functional connectivity [17] . There has been a recent surge of interest in examining the large-scale ( i . e . pair-wise connectivity throughout the whole-brain ) functional network architecture of the brain as a function of various cognitive processes or individual variation [18] . This is often done by first defining a set of functional “nodes” based on spatial ROIs and then conducting a connectivity analysis between the nodes based on their FMRI timeseries . Large-scale functional connectivity patterns have been successful in predicting age [19] as well as subject-driven mental states such as memory retrieval , silent-singing vs . mental arithmetic [20] and watching movies vs . rest [21] . It remains to be determined however , whether whole-brain connectivity can be used to decode very similar stimuli that differ by only one or a few subtle characteristics , such as the emotional expression of an unattended face . If so , then functional connections that discriminate between the two conditions can be interpreted as being uniquely related to the parameter of interest that varies across both conditions . Although multivariate pattern analyses are more sensitive than group , model-based approaches , one disadvantage is decreased interpretability and quantification of the precise relationship among features related to a certain condition [8] . However , since this approach exploits the information inherent in the joint responses of many functional connections , an advantage is that pattern classification of similar conditions coupled with feature selection and identification can be used as a means to identify condition-dependent , large-scale functional connectivity , without the need to correct for tens of thousands of multiple comparisons . This approach can be used for hypothesis generation to identify groups of functional connections associated with a condition , which can then serve as connections and regions of interest for more rigorous and mechanistically revealing approaches such as effective connectivity [22] . Here we estimate the large-scale functional networks of implicit fear processing using a blocked design and Blood Oxygen Level Dependent ( BOLD ) image acquisition , during which subjects were instructed to identify the color of pseudo-colored fearful and neutral faces ( Figure 1 ) . We applied atlas-based parcellation to derive several hundred nodes throughout the whole-brain and computed thousands of pair-wise correlations ( 40 total time points , or 80 s worth of fMRI data ) during each of two conditions: implicit processing of fearful and neutral faces . We then employed multivariate pattern analyses in conjunction with linear filter feature selection to identify functional connections whose pattern could distinguish between implicit processing of fearful and neutral faces within individual subjects , using training data from separate subjects . We plotted classification accuracy vs . number of included features to approximate the minimum number of informative features , and then identified these features ( functional connections ) on a neuroanatomical display . See Figure 2 for an outline of the analysis scheme . Our primary objective was to test the hypothesis that condition-specific , functional connectivity over the whole-brain ( here Pearson correlation using 40 time points of fMRI data per example ) contain enough information to discriminate between implicitly presented fearful and neutral faces , and to identify the functional connections that are most informative in this decoding task . A secondary objective was to compare the decoding accuracies achieved when using interactivity ( pair-wise correlations ) vs . activity ( i . e . beta estimates from SPM maps ) . We show that a small subset of connections estimated across the whole-brain can predict , or “brain-read” , implicitly presented fearful faces with high peak accuracies using training and testing data from separate subjects . We propose that this is a valuable , exploratory approach to estimate condition-dependent , large-scale functional connectivity and demonstrate that whole-brain patterns of interactivity are a sensitive and informative signature of cognitive-emotional perceptual states .
All procedures and tasks were reviewed for ethical concerns and protection of human subjects by appropriate local IRB boards prior to subject recruitment and data collection . The procedures described in this study of healthy adults have been approved by the Columbia University Morningside IRB ( #IRB-AAAA3690 , PI: Joy Hirsch ) and IRB ( #IRB5290 , PI: Myrna M . Weissman ) A total of 38 ( 19 female ) healthy volunteers ( mean age = 29 , SD = 6 . 9 ) with emmetropic or corrected-to-emmetropic vision participated in the study in accordance with institutional guidelines for research with human subjects . All subjects were screened to be free of severe psychopathology including Bipolar Disorder and Psychotic Disorders . Subjects performed a previously described task ( Etkin , Klemenhagen et al . 2004 ) which consists of color identification of fearful and neutral faces ( F and N respectively ) . Although backwardly masked ( subliminal ) fearful and neutral faces were also presented , here we discuss results based on the unmasked ( supraliminal ) conditions . Results based on comparisons of masked conditions are presented elsewhere ( manuscript in preparation ) . Stimuli: Black and white pictures of male and female faces showing fearful and neutral facial expressions were chosen from a standardized series developed by Ekman and Friesen [23] . Faces were cropped into an elliptical shape that eliminated background , hair , and jewelry cues and were oriented to maximize inter-stimulus alignment of eyes and mouths . Faces were then artificially colorized ( red , yellow , or blue ) and equalized for luminosity . For the training task , only neutral expression faces were used from an unrelated set available in the lab . These faces were also cropped and colorized as above . Each stimulus presentation involves a rapid ( 200 ms ) fixation to cue subjects to fixate at the center of the screen , followed by a 400 ms blank screen and 200 ms of face presentation . Subjects have 1200 ms to respond with a key press indicating the color of the face . Behavioral responses and reaction times were recorded . Unmasked stimuli consist of 200 ms of a fearful or neutral expression face , while backwardly masked stimuli consist of 33 ms of a fearful or neutral face , followed by 167 ms of a neutral face mask belonging to a different individual , but of the same color and gender ( see Figure 1 ) . Each epoch consists of ten trials of the same stimulus type , but randomized with respect to gender and color . The functional run has 16 epochs ( four for each stimulus type ) that are randomized for stimulus type . To avoid stimulus order effects , we used two different counterbalanced run orders . Stimuli were presented using Presentation software ( Neurobehavioral Systems , http://www . neurobs . com ) , and were triggered by the first radio frequency pulse for the functional run . The stimuli were displayed on VisuaStim XGA LCD screen goggles ( Resonance Technology , Northridge , CA ) . The screen resolution was 800×600 , with a refresh rate of 60 Hz . Prior to the functional run , subjects were trained in the color identification task using unrelated neutral face stimuli that were cropped , colorized , and presented in the same manner as the nonmasked neutral faces described above in order to avoid any learning effects during the functional run . After the functional run , subjects were shown all of the stimuli again , alerted to the presence of fearful faces , and asked to indicate whether they had seen fearful faces on masked epochs . Functional data were acquired on a 1 . 5 Tesla GE Signa MRI scanner , using a gradient-echo , T2*-weighted echoplanar imaging ( EPI ) with blood oxygen level-dependent ( BOLD ) contrast pulse sequence . Twenty-four contiguous axial slices were acquired along the AC-PC plane , with a 64×64 matrix and 20 cm field of view ( voxel size 3 . 125×3 . 125×4 mm , TR = 2000 , TE = 40 , flip angle = 60 ) . Structural data were acquired using a 3D T1-weighted spoiled gradient recalled ( SPGR ) pulse sequence with isomorphic voxels ( 1×1×mm ) in a 24 cm field of view ( 256×256 matrix , ∼186 slices , TR 34 ms , TE 3 ms ) . Functional data were preprocessed and processed in SPM8 ( Wellcome Department of Imaging Neuroscience , London , UK ) . For preprocessing , the realigned T2*-weighted volumes were slice-time corrected , spatially transformed and resampled to a standardized brain ( Montreal Neurologic Institute , 2×2×2 mm3 cube resolution ) and smoothed with a 8-mm full-width half-maximum Gaussian kernel . 1st-level regressors were created by convolving the onset of each block ( MF , MN , F and N ) with the canonical HRF with duration of 20 seconds . Additional nuisance regressors included 6 motion parameters , white matter and csf signal , which were removed prior to time-series extraction . For the current work , the same GLM analysis served three purposes: 1 ) facilitate removal of nuisance effects from time series prior to FC estimation using structurally ( atlas-based ) and functionally defined ROIs , 2 ) produce beta-estimates of each condition for classification analysis of spatial activity patterns and 3 ) functionally define ROIs ( nodes ) prior to FC calculation ( used for comparing results of structural vs . functional definition of nodes ) . Brain regions were parcellated according to bilateral versions of the Harvard-Oxford Cortical and sub-cortical atlases and the AAL atlas ( cerebellum ) and were trimmed to ensure no overlap with each other and to ensure inclusion of only voxels shared by all subjects ( Figure 3 , left panel ) . For each subject , time-series across the whole run ( 283 TRs ) were extracted using Singular Value Decomposition ( SVD ) and custom modifications to the Volumes-of-Interest ( VOI ) code within SPM8 to retain the top 2 eigenvariates from each atlas-based region . Briefly , the data matrix for each atlas-based region is defined as A , an n×p matrix , in which the n rows represent the time points , and each p column represents a voxel within an atlas-based region . The SVD theorem states:where UTU = Inxn and VTV = Ipxp ( i . e . U and V are orthogonal ) . The columns of U are the left singular vectors ( eigenvariates , or summary time courses of the region ) , S ( the same dimensions as A ) has singular values , arranged in descending order , that are proportional to total variance of data matrix explained by its corresponding eigenvariate , and is diagonal , and VT has rows that are the right singular vectors ( spatial eigenmaps , representing the loading of each voxel onto its corresponding eigenvariate ) . Here we retain the top two eigenvariates ( nodes ) from each region . For each atlas-based region , we opted to apply SVD over the entire time-series from each subject and then segment and concatenate the eigenvariates according to the conditions/comparisons of interest ( rather than segment and concatenate all the masks' voxels first and then apply SVD ) in order to maximize the total number of observations ( time points ) per region and also to avoid potentially introducing any artifact and unnatural variation caused by the splicing together of signal from disparate time points , which could possibly bias the SVD results . However , a potential disadvantage of this approach is that important sub-regions and associated eigenvariates within a particular atlas-based region could be missed due to variation in other conditions/blocks within the run that are not considered in the current work . This is an additional motivation to retain the top two eigenvariates from each atlas-based region , as opposed to just one . The above step resulted in a total of 270 nodes with an associated time course ( i . e . eigenvariates ) and spatial eigenmaps from the 135 initial atlas-based regions . Thus , each atlas-based region was comprised of two nodes . Interestingly , when extracting only one eigenvariates per region , maximum accuracy did not surpass 46% ( data not shown ) . This is possibly due to the fact that larger , atlas-based regions encompassed other functional sub-regions which were not included in the analysis . Another possible reason is that for many regions , the 1st eigenvariate may reflect artifact global or mean grey matter signal ( while white matter and csf signal were regressed out from nodes' time-series , global and mean grey matter signals were not ) , or it may reflect variation caused by other conditions/blocks within the run that were not considered in the current classification analyses ( see paradigm task description above ) , or a combination of all the above . Therefore we extracted two eigenvariates from each region . We note that this means it is likely that node 2 of a particular region shows functional connectivity that differentiates between conditions and node 1 of the same region has no differential connectivity . For clarity we therefore label each node using its Harvard-Oxford atlas label appended by either “_PC1” for the first eigenvariate and “_PC2” for the second . For display purposes , we calculated the MNI coordinates of the peak loading weight ( locations averaged across subjects ) for each eigenvariate from its associated eigenmap ( Figure 3 , right panel ) . Table S1 lists these average MNI coordinates for each node . For each subject , functional connectivity matrices ( i . e . where cell i , j contains the Pearson correlation between region i and region j ) were generated for implicit fearful ( F ) and neutral ( N ) conditions . The above time-series were segmented and concatenated according to conditions of interest ( 40 total time points per condition , incorporating a lag of 2 or 3 s from the start of each block ) before generating the correlation matrices . Fisher's R to Z transform was then applied to each correlation matrix . Finally for the binary classification of interest ( i . e . F vs . N ) , correlation matrices were demeaned with respect to the average between the two conditions in order to remove the effects of inter-subject variability . The lower diagonal of the above preprocessed correlation matrices ( 38 subjects×2 conditions total ) were then used as input features to predict viewed stimuli in subsequent pattern recognition experiments . We first tested for significant differences between the primary conditions of interest ( i . e . F>N ) while correcting for multiple comparisons ( False Discovery Rate , FDR ) . This yielded no significant results when multiple comparison correction was applied ( FDR , p<0 . 05 and 0 . 1 ) . This was not surprising , as multiple comparison correction was expected to be too conservative given the exceedingly high number of independent comparisons ( 36 , 315 ) . Support vector machines are pattern recognition methods that find functions of the data that facilitate classification [24] . During the training phase , an SVM finds the hyperplane that separates the examples in the input space according to a class label . The SVM classifier is trained by providing examples of the form <x , c> , where x represents a spatial pattern and c is the class label . In particular , x represents the fMRI data ( pattern of correlation strengths ) and c is the condition or group label ( i . e . c = 1 for F and c = −1 for N ) . Once the decision function is determined from the training data , it can be used to predict the class label of new test examples . For all binary classification tasks , we applied a linear kernel support vector machine ( SVM ) with a filtering feature selection based on t-test and leave-two-out cross validation ( LTOCV ) . There were 38 examples for each condition ( 2 from each subject , 76 total ) . During each iteration of 38 rounds of LTOCV , both examples ( 1 from each class ) from one subject were withheld from the dataset and 1 ) a 2-sample t-test was performed over the remaining training data ( N = 37 in each group ) 2 ) the features were ranked by absolute t-score and the top N were selected 3 ) these selected features were then used to predict the class of the withheld test examples during the classification stage . The full feature set for each example consisted of 36 , 315 correlations . If the classifier predicted all trials as positive or negative , the resulting accuracy would be 50% since the number of examples are equal for each class . We therefore report classification accuracy ( number of true positives and negatives over all trials ) vs . number of included features that have been ranked by their t-score . We assessed the significance of decoding results by computing the frequency in which actual values surpassed those from null distributions derived by randomly permuting class labels based on the method proposed by [25] , with the a slight modification to account for the dependence between pairs of examples from each subject . Briefly , to derive this null distribution , class labels within each pair conditions from each subject were randomly flipped with a probability of 0 . 5 over 2000 iterations for each number of included features . P-values for the peak decoding accuracies ( F vs . N: 100% , top 25 features ) were also calculated with respect to classification results when shuffling labels 10 , 000 times , and then subjected to Bonferroni correction for the number of total Top N comparisons ( in this case 20 ) . For SVM learning and classification we used the Spider v1 . 71 Matlab toolbox ( http://people . kyb . tuebingen . mpg . de/spider/ ) using all default parameters ( i . e . linear kernel SVM , regularization parameter C = 1 . Graphical neuro-anatomical connectivity maps of the top N features were displayed using Caret v5 . 61 software ( http://brainvis . wustl . edu/wiki/index . php/Caret:About ) . We note that different features could be selected during the feature selection phase of each round of cross-validation . Therefore in ranking the top 25 features , we first rank by total number of times that feature was included in each round of cross-validation , and then among these features , we sort by absolute value of the average SVM weight . Our intent is not to estimate the true accuracy of prediction given a completely new data set , but rather to test whether there exists information in the pattern of functional connections relevant to unattended emotion perception , and to approximate the optimal number of features that containing this information . We note that our approach ( plotting accuracy vs . number of top N features ) is not biased , since for each number of top N features , and for each round of leave-two-out cross validation , the top N features were selected from a training set that was completely independent from the testing set . If there is a true signal present in the data , we expect , and in the current data in general observe , that there is an initial rise in accuracy as more informative features are added to the feature set , and a dip in accuracy as less informative features ( i . e . noise ) are added to the feature set . Therefore in reporting classification results , we report the range of features at which accuracies first reach maximum accuracy-10% ( positive slope ) to which they reach maximum accuracy-10% ( negative slope ) , and also correct for multiple comparisons ( i . e . number of top N features tested ) using Bonferroni when reporting the p-value for the maximum accuracy achieved . For assessing the significance of the differences between decoding results ( i . e . FC as features vs . beta estimates ) we used the Accurate Confidence Intervals MATLAB toolbox for assessing whether the parameter p ( probability of correct prediction ) from two independent binomial distributions was significantly different ( http://www . mathworks . com/matlabcentral/fileexchange/3031-accurate-confidence-intervals ) . Briefly , these methods search for confidence intervals using an integration of the Bayesian posterior with diffuse priors to measure the confidence level of the difference between two proportions [26] . We used the code prop–diff ( x1 , n1 , x2 , n2 , delta ) , ( available from the above website ) returning Pr ( p1−p2>δ ) , where x1 , n1 , x2 , n2 , are number of correct responses and total predictions in two distributions being compared , and delta ( zero in our case ) is the null hypothesis difference between the probabilities .
The average response rate in the color discrimination task was 98% ( σ = 4 . 6% ) , mean accuracy was 97% ( σ = 3 . 5% ) , and mean reaction time was 0 . 65 s ( σ = 0 . 12 ) , indicating that subjects performed the color discrimination task as instructed . We applied atlas-based parcellation ( see Figure 2 ) and computed pair-wise correlations between 270 nodes ( derived from 135 atlas-based brain regions ) using 40 total time points of fMRI data that were segmented and concatenated from two conditions; unattended and nonmasked ( i . e . implicit ) fearful ( F ) and neutral ( N ) faces ( Figure 1 ) . This resulted in 36 , 315 total functional connections ( z-transformed Pearson correlations ) for each condition of interest ( F and N ) . We quantified the extent to which a subset of these functional connections could decode , or predict , the conditions from which they were derived by submitting them as features into a pattern classifier . We used a linear kernel Support Vector Machine ( SVM ) with a filter feature selection based on the t-score of each feature ( functional connectivity ) in each training set . Decoding accuracies for implicit fearful vs . neutral classifications ( F vs . N ) were plotted against the number of included features ( ranked in descending order by t-score ) in order to approximate the number of informative features relevant to the emotional expression of the facial stimulus . For implicit fearful vs . neutral ( F vs . N ) classification , accuracy reached 90% when learning was based on the top 15 features in each training set , a maximum of 100% ( p<0 . 002 , corrected ) at 25 features , and dipped back down to 90% at about 35 features ( Figure 4A ) . Anatomical display of the top 25 overall features that differed between F and N conditions revealed functional connections among occipital regions , middle and superior temporal gyrus , lateral and medial prefrontal regions , thalamus , cerebellum and insula ( Figure 4B–D , Table 1 ) . The connection that carried the most weight in the linear SVM classifier was between right angular gyrus and left hippocampus , which exhibited a greater correlation in the F vs . N condition ( Table 1 , F# 1 ) . To identify regions whose overall functional connectivity was greater during fear , the size of each node was made proportional to the sum of SVM weights of each of its connections . The node with the most positive functional connectivity during fear was the thalamus ( Figure 4B–D , large red sphere in center ) , which exhibited positively modulated functional connections with bilateral middle temporal gyrus and right insula . In addition to parcelating the brain and defined nodes based on an atlas , we also functionally defined nodes using two approaches 1 ) using the same 160 MNI coordinates as used in Dosenbach et . al . , 2010 [19] which were selected and defined based on separate meta-analyses of the fMRI literature , and 2 ) a biased approach based on 92 nodes ( 2 eigenvariates from each of 49 ROIs defined as 6 mm radius spheres centered at peak coordinates ) that were based on the GLM results from the same , whole dataset ( for F contrast F>N thresholded at p = 0 . 05 , k = 30 ) . For 1 ) achieved accuracies were 63–73% when using 75 to 130 features , and for 2 ) accuracies between 76–86% were obtained when using 80 to 140 features ( data not shown ) . Approach 2 ) is biased in that we defined our nodes based on the GLM results of the whole data set , and as such provides an upper bound on the expected accuracies when functionally defining nodes based on the GLM results in separate training sets during each iteration of LTOCV . Therefore we conclude that the above whole-brain , atlas-based approach , which achieved 90–100% accuracy with 15–35 features when using unbiased feature selection , is optimal to using functionally defined nodes . To compare the information content of patterns of interactivity ( i . e . functional connections used above ) vs . patterns of activity we also attempted F vs . N classifications using beta estimates , which are considered summary measures of activation in response to each condition . In order to make feature-selection/LTOCV and SVM learning more computationally tractable , preprocessed functional data were resized from 2×2×2 mm voxel resolution to 4×4×4 mm resolution , and subject-specific GLM models were re-estimated , resulting in a reduction of total feature space per example from ∼189 , 500 betas to ∼23 , 500 . Feature selection , LTOCV and SVM learning proceeded exactly as above for FC data . We observed accuracies of 66%–76% with ∼500 to 2600 features , with peak accuracy at 76% ( p = 0 . 0044 , uncorrected ) at ∼1900 features ( Figure 5A ) . The most informative voxels encompassed many distributed regions that included dorsolateral prefrontal/opercular cortex , fusiform gyrus , lateral occipital cortex , superior temporal gyrus , anterior cingulate , amygdala , parahippocampal gyrus , ventrolateral prefrontal cortex , pulvinar , precuneus , cerebellum , inferior parietal lobe and insula ( Figure 5B ) . Although significantly above chance , and despite the involvement of many more regions , maximum accuracy using betas was significantly less than the maximum accuracy achieved with FC ( 76%<100% , p = 5 . 37×10−7 ) . We performed additional classifications using betas derived from the original , smaller voxel-sizes and with the addition of an initial ( positively biased ) feature selection step over the whole-dataset for the same issues of technicality stated above . This also served to estimate an upper bound on the expected accuracy when using beta-values: if maximum accuracy achieved was still less than when using functional connectivity with unbiased feature selection , then we can more readily conclude that functional connectivity features are more “informative” than beta estimates ( when using the Canonical Hemodynamic Response Function ( HRF ) to model activation ) . For this analysis , the initial ( biased ) feature selection employed an F-test of the contrast F>N thresholded at p<0 . 01 , cluster threshold = 20 , resulting in 4 , 226 total initial features . Feature selection/LTOCV and classification again proceeded as above across the range of 1 to 4000 features . In spite of initially biased feature selection , F vs . N classification reached 92% maximum accuracy ( data not shown ) . In addition to using beta maps throughout the whole-brain , we derived beta weights using the same summary time courses ( eigenvariates ) that were extracted and used to compute pair-wise FC ( 270 total betas per condition per subject ) . For this , the GLM analysis was kept the same as above except that previously included nuisance regressors ( 6 motion , mean white and mean csf ) and a low-pass filter were not included , since they were already removed from the time courses during extraction . Resulting estimated beta weights were then used as features to predict fearful vs . neutral faces using the exact same procedure when using whole-brain FC . Accuracies of between 69–79% were achieved with between 40 to 150 features ( data not shown ) .
It is clear that fearful emotion processing and its behavioral consequences involve the complex interactions among many distributed regions [42]–[44] . Among these , the amygdala and its interactions with the frontal and visual cortex are critically involved in attended and pre-attentive threat and emotion processing [9] , [13] , [45] , [46] . Numerous previous studies have examined functional interactions between amygdala and several other regions in the fear and facial emotion processing pathway . Usually these have used Psycho-Physiological Interaction ( PPI ) analysis to study the functional connectivity of a seed region , often the amygdala , with the rest of the brain during a fearful relative to non-fear perceptual or cognitive state [12] , [46] . Other studies employed effective connectivity measures such as structural equation modeling ( SEM ) and dynamic casual modeling ( DCM ) to examine multiple interactions among a more limited set of a priori defined regions [14] , [16] . In contrast to the above-mentioned studies , the current approach is relatively model-free in that we estimate functional connectivity throughout the whole-brain without a priori restrictions based on anatomically defined areas or seed regions . We estimate network connections using simple correlation measures , similar to a previous study that demonstrated condition dependent modulations in large-scale ( 41 nodes ) functional connectivity across various syntactical language production tasks [47] , but on a much larger scale ( 270 nodes in the current analysis ) . We then identified a subset of functional connections whose pattern could discriminate between implicit fearful and neutral face processing . There is considerable interest in examining the large-scale functional network architecture of the brain as a function of various cognitive processes or individual variation [18] . This is often done by first defining a set of functional “nodes” based on spatial ROIs and then conducting a connectivity analysis between the nodes based on their FMRI timeseries . Group-based statistical parametric mapping can then be applied to resulting connections [48] . However , as the number of nodes ( N ) increases , the number of connections increases exponentially ( # connections = ( N* ( N−1 ) ) /2 ) resulting in a multiple comparisons problem , and hindering the exploration-based query of condition-specific whole-brain functional connectivity on a large-scale . The equivalent of cluster-extent thresholding for graphs has been proposed , such as the Network Based Statistic [49] , which estimates the probability of observing groups of linked , suprathreshold edges based on chance . However , inferences can only be made on groups of interconnected edges , not individual ones . In addition , there is a substantial loss of information in model-based approaches when conducting statistical inference on signals ( functional connections ) averaged over a group of subjects , and discounting the joint responses among many functional connections . Here , we present a novel alternative to identify functional connections of interest based on their information content in machine-learning based multivariate pattern analyses that attempt to discriminate between two conditions that differ based on a parameter of interest ( in this case the emotion expression of a presented face ) . For this we used linear filter feature selection and plotted classification accuracy vs . number of included features in order to determine the number of features required to distinguish between conditions , and then identified the top N features on neuroanatomical display . Large-scale functional connectivity and network analysis has been increasingly used as the tool of choice for extracting meaningful and understanding complex brain organization [17] , [18] . In the current work we applied simple Pearson correlation to estimate the large-scale functional connectivity of implicit threat-related emotion and ambiguous facial processing using a block-design . Previous work based on simulations has indicated that correlation-based methods , including Pearson correlation , are in general quite successful in capturing true network connections [18] . Here we “validated” the estimated connections by testing whether a subset of features could be used to decode ( “brain-read” ) the emotional expression of the facial stimulus that was presented during each block . For this we applied Multivariate Pattern Analyses ( MVPA ) techniques similar to those used previously to decode categories of viewed stimuli [50]–[54] , orientation [55] , [56] , and the decisions made during a near-threshold fearful face discrimination task [57] . In contrast to the above-mentioned studies , which applied MVPA to the activity of spatially distributed regions and/or voxels , in the current work we applied pattern analysis to the correlations , or interactivity , between regions distributed throughout the whole-brain . We compared the decoding accuracy when using correlations as features versus beta estimates , ( i . e . summary measures of activation amplitudes for each condition for each voxel ) . We observed that the peak classification rate when using betas ( 76% , ∼1900 features ) was significantly lower than that achieved using FC ( 100% , ∼25 features ) . Even with an additional , initial feature-selection based on the entire data set which positively biased results , peak decoding accuracies when using ∼4 , 000 beta values ( 92% ) were lower than those reached when using only ∼25 correlations as features and unbiased feature selection ( 100% ) . This suggests that there is substantially more information , relevant to cognitive-emotional neural processing , that is contained in the interactions between regions than is typically realized through standard univariate approaches . However , it should be noted that this requires enough TRs ( time-points ) to compute meaningful correlations between brain regions for a particular condition , and would thus in general be impractical for decoding single-trial or event-related data . We observed that using whole-brain , anatomically defined ROIs to define nodes for whole-brain FC estimation yielded much higher classification rates than using nodes that were functionally defined ( either from other meta-analyses or coordinates defined from GLM analysis of these same data ) . This was not too surprising , as these functionally defined ROIs were smaller ( 6 mm radius spheres centered around peak F-value coordinates from the contrast of F>N obtained from the GLM vs . atlas-based masks ) , and hence provided considerably less coverage of the brain . In addition , the GLM framework relies on multiple assumptions ( i . e . model/shape of hemodynamic response function , effects add linearly , etc . ) [58] and regions that show activation to a stimulus ( i . e . sustained increase in signal amplitude during the duration of a block ) may not necessarily exhibit differential functional connectivity and vice versa . These observations further the notion that there exists substantial information in whole-brain large-scale functional connectivity patterns , the nodes of which may not be captured or revealed adequately through standard GLM approaches . Previous simulations have raised concerns regarding the use of atlas-based approaches for parcellating the brain [18] . Because the spatial ROIs used to extract average time-series for a brain region do not likely match well the actual functional boundaries , BOLD time-series from neighboring nodes are likely mixed with each other . While this hampers the ability to detect true functional connections between neighboring regions , it has minimal effect on estimating functional connectivity between distant regions . This perhaps explains why in this study most of the functional connections that discriminated between fearful and neutral faces are long-distance . Future experiments using non-atlas based approaches would likely lead to better estimates of shorter-range functional connections . We also note that the current atlas-based approach may have under-sampled the prefrontal cortex , and that possible future improvements could break up the prefrontal regions into smaller pieces in order to sample more nodes from this area . Using Pearson correlation , it is possible that any association between two brain regions is the result of a spurious association with a third brain region . Another limitation of the current study is the required amount of data used to extract quality features of brain activity . Our use of correlations as features required a substantial number of time points ( i . e . 40 scans per condition per subject ) relative to previous studies of decoding emotion perception . Given this , it was not feasible to sample enough examples within a single or few subjects as is typical in multivariate pattern analysis studies , and we instead pooled examples across multiple subjects . On the other hand , the fact that reliable classifiers could be learned using examples from separate subjects speaks to the generalizability of our obtained results . | Brain activity is increasingly characterized by patterns of pair-wise correlations ( large-scale functional connectivity ) across the whole brain obtained from Blood Oxygen Level Dependent ( BOLD ) functional magnetic resonance imaging ( fMRI ) . Typically this is done during resting states ( i . e . no presented stimulus ) to differentiate subjects based on individual variation or diagnosis . In the current work , we identify such patterns that are a sensitive signature of unattended processing of threat-related stimuli , allowing one to “brain-read” whether an individual was presented with a neutral or fearful face while they attended to non-expression-related stimulus features . These results further the understanding of the neural mechanisms sub-serving threat-detection and facial affect processing in healthy subjects , and may also help further our understanding of various disorders , such as anxiety and autism , which exhibit anomalies in these processes . At the same time , we propose an exploratory and sensitive approach for the identification of condition-dependent , large-scale functional connectivity . This approach is not based on statistical inference on functional connections averaged across subjects and contrasted between two conditions , but rather based on the informative contribution of each functional connection when attempting to predict between two conditions , using machine-learning based multivariate pattern analysis on training data from separate subjects . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
]
| [
"psychology",
"social",
"and",
"behavioral",
"sciences",
"behavior",
"biology",
"computational",
"biology",
"neuroscience",
"neuroimaging"
]
| 2012 | Decoding Unattended Fearful Faces with Whole-Brain Correlations: An Approach to Identify Condition-Dependent Large-Scale Functional Connectivity |
In the Indian subcontinent , about 200 million people are at risk of developing visceral leishmaniasis ( VL ) . In 2005 , the governments of India , Nepal and Bangladesh started the first regional VL elimination program with the aim to reduce the annual incidence to less than 1 per 10 , 000 by 2015 . A mathematical model was developed to support this elimination program with basic quantifications of transmission , disease and intervention parameters . This model was used to predict the effects of different intervention strategies . Parameters on the natural history of Leishmania infection were estimated based on a literature review and expert opinion or drawn from a community intervention trial ( the KALANET project ) . The transmission dynamic of Leishmania donovani is rather slow , mainly due to its long incubation period and the potentially long persistence of parasites in infected humans . Cellular immunity as measured by the Leishmanin skin test ( LST ) lasts on average for roughly one year , and re-infection occurs in intervals of about two years , with variation not specified . The model suggests that transmission of L . donovani is predominantly maintained by asymptomatically infected hosts . Only patients with symptomatic disease were eligible for treatment; thus , in contrast to vector control , the treatment of cases had almost no effect on the overall intensity of transmission . Treatment of Kala-azar is necessary on the level of the individual patient but may have little effect on transmission of parasites . In contrast , vector control or exposure prophylaxis has the potential to efficiently reduce transmission of parasites . Based on these findings , control of VL should pay more attention to vector-related interventions . Cases of PKDL may appear after years and may initiate a new outbreak of disease; interventions should therefore be long enough , combined with an active case detection and include effective treatment .
VL , also called Kala-azar ( KA ) , causes more than 50 , 000 deaths each year worldwide [5] . The name ‘Kala-azar’ ( ‘black-fever’ ) originates from India; it refers to the hyperpigmentation of the skin during the course of the disease . In contrast to CL and MCL , where the parasites are localised in tegumentary tissue , VL is a systemic infection of the phagocytic and reticulo-endothelial system; this infection includes the lymph nodes , spleen and liver . Clinical symptoms of VL are prolonged fever , fatigue , weight loss , bleeding tendency and an enlarged spleen and liver . Pancytopenia is also a characteristic sign . PKDL , first described by Brahmachari in 1922 [6] , can occur as a sequel to KA . It usually develops 6 months to several years after treatment and putative recovery from KA [7] . Poor treatment compliance is thought to favour the occurrence of PKDL [8] , but a small fraction of PKDL cases have no reported history of KA [9] . The nodular lesions of PKDL patients usually contain many parasites [10] . PKDL is not a life-threatening condition , laboratory confirmation is challenging , and the treatment burdens the patients ( in Bangladesh , current treatment guidelines call for 120 intramuscular injections of sodium stibogluconate [9] ) . Therefore , it is assumed that many patients remain undiagnosed and untreated , and PKDL patients are thought to act as a reservoir of infection [11] . Cases with VL and HIV co-infection have been reported from 35 countries . Both cellular and humoral responses to Leishmania are diminished in such patients [12] , leading to an increased risk of developing VL after Leishmania infection and a higher rate of treatment failure [13] , [14] . The rate of HIV-VL coinfection in the Indian subcontinent is still low but rising in Ethiopia . HIV-coinfected patients are relevant for the study of transmission , as they are more infectious due to higher parasitaemia [15] . The classical confirmatory test for VL used to be the microscopic detection of amastigote parasites from aspirates of lymph nodes , bone marrow or spleen [16] . This invasive method has largely been replaced by antibody-detection using rK39 immunochromatographic tests or by the direct agglutination test ( DAT ) [2] . Several PCR techniques have been developed . They seem to be more sensitive for detecting asymptomatic infections than the antibody tests [17] , as it is hypothesised that PCR of peripheral blood will be the first marker to become positive after infection ( before antibody seroconversion takes place ) . The leishmanin skin test ( LST ) is a marker of the delayed-type hypersensitivity reaction and is used in epidemiological surveys to study what proportion of the population has become ‘immune’ to leishmanial infection after exposure . When taking a bloodmeal on an infected host , female sand flies ingest the amastigote form of the parasites . In the sand fly gut , the parasites develop into procyclic promastigote flagellar forms , which divide and later on differentiate into metacyclic promastigotes . These forms migrate to the pharyngeal valve and can then be transmitted at subsequent blood meals . In the human host , parasites change back to amastigotes and multiply in cells of the mononuclear phagocyte system , impeding the immune defence mechanisms of macrophages [18] . In 2005 , the governments of India , Nepal and Bangladesh started the first regional VL elimination program , aiming to reduce the annual incidence of VL to less than 1 per 10 , 000 by 2015 . The program focuses on treatment-related control strategies , such as early diagnosis and complete treatment , and vector-related control strategies , including indoor residual insecticide spraying . Currently , four drugs for VL treatment are available . 1 ) Pentavalent antimonials have been the first-line treatment for over 70 years . This treatment is long ( 20 to 30 days ) , is toxic ( 3–5% deaths due to treatment ) and is accompanied by increasing failure rates; for instance in foci of the Bihar state , India , with up to 60% treatment failures , a phenomenon that is assumed to be caused by drug resistance [19] . 2 ) Miltefosine is the first oral drug against VL and has been recommended as first-line drug in the VL elimination initiative but is teratogenic and is suspected to rapidly give rise to resistance due to its long half-life . 3 ) Amphotericin B is used in two formulations: “Conventional” Amphotericin B and “Liposomal” Amphotericin B ( Ambisome ) . Treatment with the conventional formulation is long ( 30 days ) and can have life-threatening side effects whereas treatment with the liposomal formulation is shorter , has similar high efficacy and fewer side effects; Yet it is too expensive and complex in its administration for large-scale application in developing countries . However recently , a single-dose Ambisome schedule was found highly effective in India [20] . 4 ) Paromomycin ( PMM ) was registered in 2006 in India and is currently being tested in a phase IV trial [2] , [21] . The first mathematical study of the dynamics of KA employed a deterministic model to explain the observed inter-epidemic periods between 1875 and 1950 in Assam , India [25] . This model was later extended to canine visceral leishmaniasis in Malta , considering three different serological tests for the presence of specific antibody in dogs [26] , [27] . In a further extension , the efficacies of various control methods , such as vaccination , killing infected dogs , drugs and insecticides , were investigated by means of sensitivity analyses [28] . A recent modelling paper addressed the question of under-reporting of VL in Bihar , India . [29] . Our model focuses on the transmission dynamics of L . donovani in the Indian subcontinent and extends the before-mentioned modelling findings with ( i ) quantifications of epidemiological parameters for the natural history of infection and transmission dynamics of L . donovani and ( ii ) databased predictions for the effects of different intervention strategies; in particular , the roles of asymptomatic and symptomatic infections are considered .
The transmission dynamics of L . donovani in the Indian subcontinent was modelled deterministically by a system of ordinary differential equations . Figure 1 describes the model graphically , and Table 1 , 2 and 3 provide lists of parameters and variables with references . In this section , we describe the basic structure of the model; the full model , considering also possible animal hosts and immuno-compromised humans , is provided in Text S1 with Figure S1 showing the full model . The basic Susceptible-Infected-Recovered ( SIR ) type model was modified with respect to the history of infection , considering five stages in the natural history of infection . These stages were described based on different combinations of three diagnostic markers to categorise people living in endemic areas: PCR ( index P , earliest marker for L . donovani infection ) , DAT ( index D , antibody response ) and LST ( index C , suggesting a state of ‘cellular’ immunity ) . Superscripts “+” and “−” represent positive and negative results , respectively . If the infection proceeded asymptomatically , we assumed that the human host ( indicated by subscript H ) typically passed through the following five stages: SH: Susceptible stage ( P−D−C− ) . Hosts are negative for all three markers and can become infected in the future . IHP: Early asymptomatic , infectious stage ( P+D−C− ) . The parasite can be detected by PCR , but there is not yet any humoral or cellular response in the host . IHD: Late asymptomatic , infectious stage ( P+D+C− ) . Hosts are still PCR-positive and antibodies can be detected by DAT . RHD: Early recovered stage ( P−D+C− ) . The parasite cannot be detected anymore , but hosts are DAT-positive and not yet LST-positive . RHC: Late recovered stage ( P−D−C+ ) . Hosts are DAT-negative but still LST-positive and assumed to be protected against re-infection . Model solutions were numerically computed using the software Matlab Version 7 . 120635 ( R2011a ) by a Runge-Kutta algorithm with variable step size ( procedure ode15s ) . Parameters were estimated by fitting the model to observations from the KalaNet project ( DAT- and PCR-positivity , symptomatic VL and sand fly infection ) and information from the literature ( LST-positivity , prevalence of HIV and prevalence of HIV-infected among symptomatic VL ) . We used the model to estimate the following eight parameters: The parameter vector was estimated by Maximum Likelihood with likelihood functionwhere by exponents k1 to k22 represent sample sizes observed in the KalaNet trial or taken from the literature ( for references see Table 1 , 2 , 3 and Table S1 ) . These exponents are partially not integer values , because sample sizes for the estimation had to be back calculated considering the prevalence of HIV-positive hosts , yielding The likelihood was maximized by minimizing the negative log-Likelihood −ln ( L ) , using Matlab procedure fminsearch . Confidence intervals were computed by the profile likelihood [34] , i . e . the parameter of interest was minimised ( lower confidence limit ) or maximised ( upper confidence limit ) conditional on that the likelihood worsens only according to χ2/2 with 1 degree of freedom , i . e . 3 . 84/2 , whereby the other 7 parameters may vary as nuisance parameters . Parameter values , estimates and confidence intervals are shown in Tables 1 , 2 , 3 and Table S1 .
The parameter estimates ( Table 1 , Table 2 , Table 3 and Table S1 ) originate from fitting the model to data of which prevalences have been estimated as described in the Methods section were as follows 26% of the human population ( SH ) were negative for all three markers of PCR , DAT and LST ( P−D−C− ) , 10% ( IHP ) were PCR-positive ( P+D−C− ) , 2% ( IHD+IHS+IHT1+IHT2+IHL ) were PCR- and DAT-positive ( P+D+C− ) , 12% ( RHD+RHT+RHL ) were DAT-positive ( P−D+C− ) and 50% ( RHC ) were LST-positive ( P−D−C+ ) . Model-based rates and durations in the natural history of infection were estimated from the prevalences for humans as follows . Humans remained on average 1/ ( λH+μH ) = 150 days in the susceptible state ( SH ) before they became infected . After infection , PCR-positivity lasted on average for 1/ ( γHP+μH ) +1/ ( γHD+μH ) = 72 days ( approx . 95% CI: 69 to 75 days ) and overlapped with DAT-positivity , which lasts for asymptomatic individuals for 1/ ( γHD+μH ) +1/ ( ρHD+μH ) = 86 days ( approx . 95% CI: 74 to 99 days ) . For an overlap period of about 1/ ( γHD+μH ) = 12 days , asymptomatic individuals were positive for both markers ( late asymptomatic state IHD ) . This means that individuals without symptomatic disease need on average about 146 days to develop LST-positivity after a PCR-positive finding . LST-positivity lasted on average 1/ ( ρHC+μH ) = 300 days ( approx . 95% CI: 255 to 348 days ) . As population turnover alone is not sufficient to yield a prevalence of 26% susceptibles ( negative for all three markers: P−D−C− ) , loss of LST-positivity must be assumed to explain these , together with a prevalence of 50% LST-positive individuals in the population . Taken these durations together , humans with an asymptomatic course were infected on average every 596 days ( = 150 days SH+60 days IHP+12 days IHD+74 days RHD+300 days RHC ) . As heterogeneities are not taken into account , one might crudely assume that humans become infected every one to three years . Given a 0 . 5% prevalence of infected sand flies [35] and the above-mentioned prevalences , the probability pF2 that a sand fly is infected during a blood meal on an asymptomatic host ( IHD ) was estimated at 2 . 5% with 95% CI ( 1 . 2% to 3 . 8% ) . This estimate is robust against changes in the infection probability for flies feeding on symptomatic hosts , which was assumed to be pF3 = pF4 = 100% . A sensitivity analysis in which pF3 and pF4 were reduced to 10% had no relevant effect on the estimate for pF2 , which then decreased from 2 . 5% to 2 . 3% . The reason for this minor contribution of symptomatically infected hosts to the overall transmission is simple: their prevalence is too low compared to the abundance of asymptomatically infected hosts . The estimates reported so far are correlated with the time between blood meals of sand flies ( assumed with 1/β = 4 days [36] ) and the ratio NF∶NH of vectors to humans ( estimated at 527∶100 , with 95% CINF ( 347 to 990 ) ) . We investigated interventions recommended by the VL elimination program , including treatment-related control strategies , such as early case detection or treatment optimisation . We also considered vector-related control strategies , such as breeding site control , indoor spraying or use of bed nets , to explore possibilities and constraints in VL control .
The transmission dynamics of L . donovani are rather slow , mainly due to the long incubation period and the potentially long persistence of parasites in infected humans . Parasite DNA can be detected in infected human hosts for two to three months; after seroconversion , antibodies can be detected for about the same period of time in those humans who do not evolve to clinical disease . These periods overlap for a few weeks only , when both parasites and antibodies are detectable in peripheral blood . Few infected humans become clinically sick . The majority of infected hosts are estimated to become LST-positive five to six months after infection , without showing any sign of disease . LST-positivity , which is assumed to represent a state of protective cellular immunity , is estimated to persist for about one year . Without loss of LST-positivity , the model would lead under realistic birth and death rates of humans to over 90% LST-positive individuals in the population . For analysing the process of human recovery , published studies involving the LST were used to estimate the duration of the cellular immune response based on LST-positivity . Previous studies reported prevalences of positive LST results among recovered KA patients ( who are expected to be LST-positive ) ranging from 30% to 80% [33] . These studies suggested that the lowness and variability of LST sensitivity is affected by the leishmanin antigen suspension chosen . Because an antigen based on South Asian L . donovani is not available , all LST data refer to L . infantum or L . major antigens . There is extensive cross-reactivity of patient responses to heterologous Leishmania species , but as the variable results underscore , there is a need for better standardisation and documentation of sensitivity and stability of leishmanin antigens to obtain more reliable data on the state of cellular immunity . Effects associated with differing proportions of LST-positive individuals are explored in the section on sensitivity analyses below . Apart from problems with the antigen ( see above ) , there are in principle two possibilities to explain a proportion of only 50% LST-positive individuals ( cf . before: >90% LST-positive individuals would be expected under the assumption of life-long LST-positivity ) . These two explanations depend on how infection spreads , as follows: 1 ) If infection is clustered ( e . g . , within-household transmission predominates due to a low rate of transmission between households ) then life-long LST-positivity may exist on the level of the individual , and a low , ‘average’ LST-prevalence in the population must result from the demographic process , i . e . , it must result from deaths of LST-positive individuals and the births of individuals who are LST-negative . 2 ) If infection is not clustered , but spreads homogeneously within the human population , then a low LST-prevalence in the population must involve loss of the ‘individual’ LST-positivity ( as opposed to option 1 where the ‘average’ LST-positivity gets lost ) . As the birth rate is not sufficient to explain a prevalence of 50% LST-positive individuals ( see above ) , we proceeded with the assumption of loss of LST-positivity . Under this assumption , humans in active L . donovani transmission foci were infected every one to three years . A considerable loss of LST positivity has been reported from an endemic VL focus in Ethiopia [37] . There , the authors concluded that the presumption of a life-long positive LST reaction may only hold true under circumstances of continued exposure to infected sand flies or subclinical infection . A short and effective treatment regimen or early case detection can reduce the prevalence of KA but has almost no effect on the incidence of the disease ( Fig . 2 and Table 4 ) . Treatment is thus an intervention that benefits the individual but not the population . The reason why treatment is not a measure to control the transmission of infection must be attributed to many asymptomatically infected hosts who are - despite a low infectivity for flies - responsible for the majority of parasite transmissions from humans to sand flies . As demonstrated by the sensitivity analysis shown in Fig . 2 and Table 4 ( scenario 2 ) , a beneficial effect of treatment on transmission cannot even be expected under over-optimistic assumptions; for instance , minimal duration of treatment and time until treatment start with maximum treatment efficacy . The KA elimination program aims for less than 1 case per 10 , 000 . Even under over-optimistic treatment assumptions , this target would not be reached; the model suggests that elimination of VL would not succeed if intervention was based on treatment alone . The contribution of asymptomatic infections on transmission may , however , be region and strain specific , as apparent from ratios of asymptomatic infections to KA cases ranging from 6∶1 in Brazil to 50∶1 in Spain [38] . The current vector-related interventions comprise irregular indoor spraying and exposure prophylaxis , such as untreated bed nets . As summarised by [30] , indoor insecticide spraying during the Indian National Malaria Eradication Program between 1958 and 1970 had a drastic impact on transmission of L . donovani . No VL cases were reported from the state of Bihar during that period . However , within months after the program was terminated , the first cases of KA re-appeared and at the end of 1970 , a VL epidemic struck Bihar . In 1992 , another eradication program with indoor residual DDT spraying resulted in a sharp decline from 1993 until 1999 . Whereby treatment-related interventions only reduce symptomatic infection , vector control is also efficient against asymptomatic infection ( Fig . 3 and Table 5 ) . The model predicted that VL may be eliminated by reducing the vector density by 80% , which seems to provide serious options for the VL elimination program ( results not shown ) . Prospects of successful elimination can also be expressed by the moderate basic reproduction number which this model predicts with R0 = 3 . 94 ( see Text S1 ) . As found during the KalaNet study , the use of treated bed nets may not provide this efficacy . Indoor density of P . argentipes is reported in the study to be reduced by 25% in villages that used treated nets compared with control villages . This intervention reduced the annual incidence of VL at best to 18 . 8 cases per 10 , 000 , which is far from the elimination target of less than one case per 10 , 000 . The authors hypothesised that a substantial fraction of L . donovani transmission occurring outside the house may explain the results obtained in the trial [24] . However , two other studies in the Indian subcontinent showed a much higher reduction in the sand fly density by means of vector-related interventions [22] , [39] . Despite the fact that vector control has the potential to effectively reduce L . donovani transmission , it must be complemented by the treatment of KA and PKDL cases . The latter can substantially maintain a reservoir of infection , which can initiate a new outbreak of disease after vector control is stopped . As the natural history of L . donovani transmission is a rather slow process where parasites can persist in humans over a long period , interventions will show an effect after years , not months , as demonstrated in Fig . 4 . PKDL cases may occur after years , and their final recovery may take months or years . If vector control is stopped before parasites are eliminated in the human hosts , then emerging PKDL cases can initiate a new outbreak of disease as soon as there are enough sand flies available for transmission . Thus , active case detection and effective short-course treatment of PKDL combined with effective vector control are required for VL elimination . To determine the role of LST-positive individuals ( RHC ) , the default assumption of 50% LST-positive individuals was varied between 30% and 70% ( 76% is the maximum possible value because 24% of the human population belong to other diagnostic categories , see above ) . A low equilibrium prevalence of LST-positive individuals results from a lower infection rate ( or lower values of correlated parameters , such as the biting rate or the number of vectors ) , and a high prevalence of LST-positive individuals can result from a higher infection rate ( or higher values of correlated parameters , such as the biting rate or the number of vectors ) . The only parameter estimate that was affected within the natural history of infection by this variation of the prevalence of LST-positive individuals was the rate of loss of cellular immunity ( ρHC ) . The estimate of 1/ρHC = 307 days of LST-positivity with the 95% confidence interval between 260 and 356 days ( see Table 2 ) does not support the assumption of a life-long cellular immunity based on life-long LST-positivity ( in technical terms , life-long LST-positivity would lead to an excessive accumulation of individuals in state RHC , which is not consistent with the data ) . The effects of different prevalences of infected sand flies were also analysed , as the prevalence of infected sand flies may vary regionally . The default assumption of IF = 0 . 5% infected sand flies is based on data from Nepal , whereas higher prevalences have been observed in India . A higher prevalence of infected sand flies results from a lower number of flies ( NF ) or from a higher infection probability originating from asymptomatic hosts ( pF2 ) . We assumed that the period of DAT-positivity is the same for symptomatic and for asymptomatic cases ( 1/ρHD = 1/ρHT = 74 days ) . The estimate of ρHD ( DAT-positivity in asymptomatic cases ) is robust against the assumption that DAT-positivity may last longer in symptomatic cases [40] . In that investigation antibody persistence in symptomatic cases was suggested to last on average about 4 years . Such an increase , however , can be completely compensated in the model by reducing the period of DAT-positivity in asymptomatic cases from 74 to 69 days , showing that the model behaves robust against changes in these assumptions . A full factorial sensitivity analysis is provided in Fig . S2 in the Supplement . Our simulation results show that transmission of L . donovani is predominantly driven by asymptomatically infected hosts who are not eligible for treatment . Treatment can reduce the prevalence of symptomatic disease , but the incidence of KA remains on similar levels because of an unchanged intensity of transmission . In contrast to treatment-related interventions , vector-related interventions have the potential to reduce the prevalence of asymptomatic infections and thus are the intervention of choice from an epidemiological perspective . Vector control , however , should be combined with treatment , as PKDL cases can act as reservoirs of infection . This reservoir function originates from the long period of nearly two years on average during which putatively recovered KA patients develop PKDL . | Visceral Leishmaniasis is a neglected , life-threatening disease affecting the poorest of the poor . It has received more attention in light of the regional VL elimination program . A deterministic compartmental model was developed to estimate parameters for L . donovani transmission and to optimise intervention success . Our simulation results show that transmission of L . donovani is predominantly driven by asymptomatically infected hosts who are not eligible for treatment . Treatment can reduce the prevalence of symptomatic disease , but the incidence of KA remains on similar levels because of an unchanged intensity of transmission . In contrast to treatment-related interventions , vector-related interventions have the potential to reduce the prevalence of asymptomatic infections and thus are the intervention of choice from an epidemiological perspective . Vector control , however , should be combined with treatment , as PKDL cases can act as reservoirs of infection . This reservoir function originates from the long period of nearly two years on average during which putatively recovered KA patients develop PKDL . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
]
| [
"medicine",
"public",
"health",
"and",
"epidemiology",
"infectious",
"disease",
"epidemiology",
"population",
"dynamics",
"parasitic",
"diseases",
"neglected",
"tropical",
"diseases",
"population",
"biology",
"infectious",
"disease",
"control",
"infectious",
"diseases",
"epidemiology",
"biology",
"infectious",
"disease",
"modeling",
"disease",
"dynamics",
"leishmaniasis"
]
| 2011 | Visceral Leishmaniasis in the Indian Subcontinent: Modelling Epidemiology and Control |
Studying a gene’s regulatory mechanisms is a tedious process that involves identification of candidate regulators by transcription factor ( TF ) knockout or over-expression experiments , delineation of enhancers by reporter assays , and demonstration of direct TF influence by site mutagenesis , among other approaches . Such experiments are often chosen based on the biologist’s intuition , from several testable hypotheses . We pursue the goal of making this process systematic by using ideas from information theory to reason about experiments in gene regulation , in the hope of ultimately enabling rigorous experiment design strategies . For this , we make use of a state-of-the-art mathematical model of gene expression , which provides a way to formalize our current knowledge of cis- as well as trans- regulatory mechanisms of a gene . Ambiguities in such knowledge can be expressed as uncertainties in the model , which we capture formally by building an ensemble of plausible models that fit the existing data and defining a probability distribution over the ensemble . We then characterize the impact of a new experiment on our understanding of the gene’s regulation based on how the ensemble of plausible models and its probability distribution changes when challenged with results from that experiment . This allows us to assess the ‘value’ of the experiment retroactively as the reduction in entropy of the distribution ( information gain ) resulting from the experiment’s results . We fully formalize this novel approach to reasoning about gene regulation experiments and use it to evaluate a variety of perturbation experiments on two developmental genes of D . melanogaster . We also provide objective and ‘biologist-friendly’ descriptions of the information gained from each such experiment . The rigorously defined information theoretic approaches presented here can be used in the future to formulate systematic strategies for experiment design pertaining to studies of gene regulatory mechanisms .
Cellular processes are determined by the response of regulatory sequences in DNA to signals from specific proteins called transcription factors ( TFs ) , leading to up- or down-regulation of gene expression [1] . A major class of regulatory sequences is that of cis-regulatory modules ( CRMs , also called enhancers ) : these are regions of DNA , about 500–2000 base pairs long , harboring TF binding sites that control the transcriptional levels of nearby genes . Variation of the DNA sequence in CRMs can affect gene expression , and has been linked to developmental defects and disease [2] . Even minor variations , such as single nucleotide polymorphisms ( SNPs ) , in CRMs can have significant functional impact , such as problems in fetal development [3] . Our ability to predict the impact of non-coding sequence variations on gene expression is very limited , in part due to the complexity of CRMs , and in part because such impact depends not only on the sequence itself but also the abundance and activities of relevant TFs in the cellular conditions of interest . Statistical methods based on correlations among diverse data types such as TF-ChIP , histone modifications , gene expression , etc . can reveal salient properties of CRMs such as their tissue-specific activities [4] and their major regulators [TFs] [5 , 6] , and can in some cases predict the effect of removing a TF’s influence on a CRM or gene [7–9] . Statistical and machine learning methods have recently been developed that can to some extent predict the effects of single nucleotide mutations on TF binding levels , DNA accessibility [10 , 11] , and even gene expression [12] , but these are typically not amenable to mechanistic interpretations , and are in a relatively early stage of exploration . On the other hand , biophysical models based on equilibrium thermodynamics that explicitly incorporate key interactions among TFs , DNA and the transcriptional machinery have proven powerful for mechanistic understanding of the gene regulation process [13] . Thermodynamics-based modeling of gene expression reveals the precise mapping between CRM sequence and the associated gene expression in a variety of cellular contexts , the so called ‘readout’ of the CRM . These models provide a means to formalize our assumptions about a CRM’s cis-regulatory logic , especially how its functional elements combine to regulate a transcriptional output [14–19] . They can generate predictions that can be empirically tested [20] , e . g . , by targeted misexpression or mutagenesis experiments . Indeed , they have been used to predict effects of site mutations [21] , and also promise to provide precise , mechanistically grounded predictions of the effect of minor sequence changes in CRMs [20] . Furthermore , these models can reveal ambiguities in our mechanistic knowledge about a system given existing data; pinpointing these ambiguities helps with choosing the future experiments that would best improve knowledge of the system . The success of thermodynamic models has been demonstrated in the context of systems with high-resolution gene expression measurements , such as early-stage Drosophila embryonic development [15–17 , 22] . Mechanisms influencing the precise function of a regulatory system include the number , accessibility , affinities and relative arrangement of TF binding sites within a CRM , as well as cellular concentrations of the TF molecules , and protein-protein interactions; all of these mechanisms affect the rate of transcription of the gene [23] . Thermodynamic models of CRM function encode these mechanistic factors in their parameters , which correspond to biochemical properties of the molecules controlling the gene expression . These parameters are typically computationally optimized to be assigned values that can best explain the gene expression patterns attributable to a set of CRMs [16 , 22] . When used to investigate the regulatory function of a single CRM , the thermodynamic modeling approach faces a significant challenge: non-uniqueness of the optimal models . For instance , a CRM mediating control by five or more TFs will include 10 or more free parameters in the GEMSTAT thermodynamic model [22] and we have shown previously that parameter training will converge to one of many local optima [21] . Each optimal model explains the data equally well , but uses parameters that correspond to significantly varying , often mutually incompatible mechanistic hypotheses [24 , 25] . For example , consider a gene regulated by a CRM that is under the control of two activators . Assume there are two models that explain the wild-type expression pattern . One model predicts the correct expression by using ( assigning function to ) only one activator , while the other model uses both activators . In the absence of additional biological experiments that confirm the role of each activator , both models are equally plausible . Problems arise when we try to predict , using the model , the effect of knocking down an activator or mutating its binding site ( s ) . Depending on which model we use , the predicted effect of the perturbation experiment is different: a model that does not use an activator will not predict a change due to removal of that activator’s influence . However , there is no reason to prefer one model’s prediction over the other , and the biology remains ambiguous until a new experiment is performed . We believe it is important to respect this ambiguity of knowledge when modeling gene expression data and making predictions about future experiments . In agreement with the above proposal , Samee et al . [21] laid out a new paradigm of gene expression modeling where one searches within the model’s parameter space for as many optima as possible , resulting in an “ensemble” of optimal models . ( Henceforth , different assignment of values to the model’s tunable parameters will be considered as different models . ) Each model in the ensemble is a hypothesis about the cis-regulatory mechanisms encoded in the CRM , and is also capable of making specific predictions about perturbation experiments . A simple approach to working with an ensemble of models is to make predictions by uniformly aggregating predictions of its member models . It has been shown that this “wisdom of crowds” approach can be effective: aggregated votes of many models can predict the effect of site mutations more accurately than any individual model [21] . We noticed , however , that a typical ensemble of sequence-to-expression models , e . g . , that created by Samee et al . [21] in modeling embryonic expression of the Drosophila gene ind , is not uniformly distributed in the parameter space . Rather , they are clustered in the parameter space ( Fig 1D ) , with models within a cluster predicting similar effects for a particular perturbation , but each cluster’s consensus predictions being qualitatively different from those of other clusters . Different clusters can have different ‘spans’ , i . e . , the extent to which models in that cluster differ from each other quantitatively ( in parameter values ) while producing equally good fits to available data and essentially the same predictions for future experiments . For instance , the cluster at the bottom of Fig 1D has greater span than the cluster on the top-left , which is relatively tight . The span of a cluster pertains to parameter sensitivity [24 , 26] in that region of parameter space . Furthermore , different clusters may have different representation ( number of models ) in the ensemble , and the number of represented models may not correlate with the span of the cluster . This is because we do not make strong assumptions about how the ensemble of models was obtained , beyond that it is a collection of models that fit the available data and may be located in different regions of parameter space . With these observations about ensembles of models , we sought the most appropriate way to use ensembles for making predictions and for designing future experiments . We describe here one such procedure that we developed and implemented , which allows us to make predictions with ensembles of models , and also offers a principled approach to experiment design in gene regulation studies . Briefly speaking , our modeling approach involves ( 1 ) creating a large ensemble of models that fit the available data accurately , following the sampling and optimization strategy of Samee et al [21] and ( 2 ) defining a probability distribution over the parameter space such that the ensemble of models represents regions of high probability and where each cluster of models ( roughly speaking , a distinct mechanistic hypothesis ) has approximately the same total probability as other clusters . This distribution provides a principled way for us to make aggregated predictions about any particular perturbation experiment , and to describe the uncertainty in such predictions . Additionally , we show how to measure the entropy of this probability distribution , thereby quantifying the uncertainty in parameter space [27] that remains after fitting the models to available data . Noting that the ensemble of models consistent with available data changes ( typically shrinks ) upon performing an additional experiment , we suggest that the difference of entropies of the probability distributions before and after an experiment ( i . e . , information gain ) may be used to score the ‘value’ of the experiment . We can use this value as a score to compare different experiments , the experiment with greater score being deemed the more informative experiment . The ability to assign information theoretically-grounded ‘values’ to experimental results is significant , since it paves the way for principled experiment design [28 , 29]
We consider the class of mathematical models that predict the gene expression level driven by a cis-regulatory module ( CRM ) from the latter’s sequence , given prior knowledge of relevant transcription factors ( TFs ) , their in vitro DNA-binding affinities ( motifs ) , and their concentration levels in the cellular context of interest . Several such models have been investigated in the literature [15–17 , 20] , and we work with the GEMSTAT model [22] , which we developed previously and which we are most familiar with . The GEMSTAT model has two free ( tunable ) parameters for each relevant TF , one corresponding to its binding strength for the consensus site and one corresponding to its potency as an activator or repressor . The model also has optional free parameters for any TF-TF cooperative interactions that the modeler may choose to include . Assigning values to these free parameters specifies a model completely , allowing it to predict gene expression in any cellular context where TF concentrations are known . Typically , optimization strategies are used to identify the parameter setting ( s ) that accurately predict gene expression driven by a CRM in multiple cellular contexts [30] . In light of the observations made in Introduction , we sought to first construct an ensemble of models that are widely spread in parameter space , and thus represent different mechanistic explanations of data . A model is included if its goodness-of-fit score–sum of squared errors or ‘SSE’ between known and predicted expression levels in multiple cellular conditions–is below a threshold . We noted that the number of TFs in common modeling scenarios is less than 10 [15–17 , 20 , 22] , and the number of free parameters in the range of 10–20 . This led us to consider uniform sampling of the parameter space as the first step of ensemble construction . We followed the approach of Samee et al . [21] ( Fig 1A–1D ) , performing extensive uniform sampling from the space ( millions of samples ) , followed by filtering of promising models ( SSE score below a modest threshold ) , local optimization seeded by these promising models , and a final round of filtering on the optimized models ( SSE score below a strict threshold ) . ( See Methods for details . ) This procedure allows us to construct a large ensemble of models representing many or all optimal regions of the parameter space . We provide more details of ensemble size and composition later , in the context of specific gene models . An ensemble of models can be used to make predictions by aggregating ( averaging ) the predictions made by each member model . However , this approach ignores the fact that the ensemble construction ( outlined above or by a similar method ) likely results in some regions of parameter space being over-represented in the ensemble . Models belonging to the same region , i . e . , proximal to each other in the parameter space , are presumed to represent qualitatively similar mechanisms of CRM function . Thus , the ensemble’s aggregate predictions may be biased towards one or a few mechanistic hypotheses . We therefore sought a more nuanced way to aggregate model predictions , by defining a probability distribution over parameter space that captures how the fit models are spread across different regions of the space but discounts for unequal representations of ( number of models in ) different regions . Such a probability distribution can then be used to make predictions about new experiments and also to score the uncertainty of mechanistic explanations offered by the ensemble . We also note that constructing this distribution has close ties to the kernel density estimation problem [31] but is different because the ensemble is not a collection of IID samples drawn from the desired population . The simplest distribution to consider is a discrete uniform distribution over the models in the ensemble; e . g . , Fig 1E shows such a distribution over an ensemble of four models in a toy 1-dimensional parameter space . In the continuous parameter space , highly proximal models are likely to have similar goodness-of-fit , therefore we smoothen the discrete distribution by centering a Gaussian distribution at each model in the ensemble and constructing a uniform mixture ( Fig 1F ) . This mixture of Gaussian distributions provides a continuous distribution , but if one region of the space is over-sampled in the ensemble , the distribution puts undue weight in that region; e . g . , the three closely-related models on the left in Fig 1F together carry about three times the probability mass as that around the isolated model on the right . In light of this observation , we first cluster models , each cluster roughly corresponding to a distinct mechanistic hypothesis , and define the overall probability distribution to be a mixture of distributions representing each cluster . Since we lack any additional knowledge to prefer one cluster over another , we assume uniform mixture weights for the clusters . The probability distribution representing each cluster , in turn , is a mixture of Gaussian distributions whose means are the models in that cluster . Thus , Fig 1G shows a mixture of two distributions ( red and blue ) representing the two clusters , with the red distribution in turn being a uniform mixture of Gaussians centered on the three models in that cluster . Fig 1H shows a similar construction , now for a 2D parameter space , beginning with the given filtered ensemble ( θ1 , θ2 , … θM ) , identifying three clusters and constructing the mixture probability distribution . ( For more details , especially the construction of covariance matrices for these distributions , see Methods . ) The probability distribution over models , constructed as above , can be used in the following ways: Sequence-to-expression models enable us to propose mechanisms for gene expression regulation , that may then be confirmed by performing perturbation experiments such as TF knockout or site mutagenesis , followed by expression assays that inform us about how the gene expression changes in the perturbation condition . Some experiments result in greater gene expression changes than others , and it is natural to want to characterize ‘what was learned’ from each experiment , as well as quantify how informative that experiment was . Here , we demonstrate such an exercise in systematic experimentation in the gene regulation context , using the ensemble modeling framework described above . Our first set of demonstrations are in the context of the regulatory mechanisms of an early development gene in D . melanogaster—the intermediate neuroblasts defective ( ind ) gene . We chose this gene because it is known to be regulated by a well-defined enhancer , and its major regulatory inputs are well characterized . The gene was characterized by Weiss et al . [32] and Stathopoulos and Levine [33] , among others , and was the subject of systematic modeling by Samee et al . [21] . It is expressed in a lateral stripe along the dorso-ventral axis of the early embryo ( S1A Fig , black curve ) , with activation from the TFs Dorsal ( DL ) and Zelda ( ZLD ) , and repression by the TFs Snail ( SNA ) , Ventral nervous system defective ( VND ) and Capicua ( CIC ) ( S1A Fig ) . In addition to the wild-type expression pattern of this gene , its expression has been experimentally recorded under several perturbation conditions ( S1 Table ) , surveyed by Samee et . al [21] and further discussed below . Despite the knowledge of a fairly complete set of regulatory inputs , several ambiguities remain about the cis-regulatory logic of the ind enhancer . This is evident when we construct an ensemble of models that predict the known expression pattern of ind from its enhancer sequence along with TF concentration profiles along the D/V axis . S1B Fig shows that the ensemble’s mean prediction ( magenta curve ) for these wild-type conditions fits the wild-type expression profile accurately , and with little variation among different models ( pink curves are models in the ensemble ) but S1C Fig reveals that most of the 13 parameters of the model exhibit substantial variability , a point also illustrated by the marginal distributions of ten of the parameters ( S1D–S1F Fig ) . The high degree of uncertainty is not surprising , given that data from only one experiment–the wild type condition–for a single enhancer was used to train the ensemble . It also means that results of various perturbation experiments may prove informative about this gene’s regulatory mechanisms , an avenue that we pursue next . First , we worked with a ‘synthetic true model’ MST that allows us to predict results of various perturbation ‘experiments’ in silico . This synthetic true model MST was carefully chosen from among the ensemble of models consistent with wild-type data , described above . ( See S3 and S4 Figs for details . ) We used MST to individually predict the effects of ( a ) each TF’s knockout and ( b ) removing the strongest site of each TF in the enhancer , and treated these predicted gene expression patterns ( Fig 2A , green curves ) as the ‘true’ results of those hypothetical or ‘in silico perturbation experiments’ . We used each of the 10 in silico experiments to construct a ‘filtered ensemble’ ( average predictions shown in Fig 2A , magenta curves ) , computed its entropy score , and thus assigned an information theoretic ‘value’ to the experiment ( Fig 2B ) . We noted that the magnitude of change in the expression profile resulting from a perturbation experiment does not necessarily reflect the value of the experiment . For instance , it is possible to obtain new information from a perturbation experiment where the expression pattern remains unchanged from wild-type , a case in point being the SNA knockout experiment ( Fig 2A ) , with assigned value 1 . 66 –apparently many models consistent with wild-type data cannot explain this experiment and are removed in the filtered ensemble from its results . Conversely , an experiment with a more substantial expression profile change may not add anything to our knowledge of the regulatory mechanism . For instance , the DL knockout experiment shows peak ind expression diminishing by ~60% ( Fig 2A ) but is assigned a value of 0 . 32 ( Fig 2B ) , among the lowest of the 10 experiments; this is because most models capable of explaining the wild-type ind pattern apparently use DL as activator , so knocking out DL does not provide much new information . The same is not true of the experiment where the strongest DL site is removed , an experiment with minor impact on expression ( Fig 2A ) but a relatively high assigned value of 1 . 34 . This points out that even if the involvement of a TF is beyond doubt , there may be uncertainty regarding the strength of its regulatory input and the mediatory role of each of its binding sites . We noted ( Fig 2B ) the same trend—that the value of strongest site mutagenesis is greater than that of TF knockout–for the other activator ( ZLD ) . On the other hand , for perturbations involving repressors ( SNA , VND , CIC ) the value of the site mutagenesis experiment is less than that of TF knockout in all three cases . Also , for comparison , we show in Fig 2C the relative values of the 10 ‘experiments’ under a more simplistic scheme that evaluates each experiment by the reduction in entropy assuming a discrete uniform distribution on all models in an ensemble . We note that the two schemes largely agree with each other in this evaluation , though this may not be true in general , depending on how an ensemble of models is generated . Finally , we note that the observations above were made with a specific choice of the ‘synthetic true model’ MST , that furnished ‘experimental’ results , but the reported trends , e . g . , large information gain from a perturbation experiment with little effect on expression , or little gain from an experiment with large effect , were unchanged when we repeated the entire exercise with a different choice of MST ( S5 and S3B Figs ) . In this section , we will examine results of real perturbation experiments pertaining to the ind gene reported in the literature and evaluate each experiment in the way described above . In addition to the wild type gene expression pattern of the ind gene ( S1A Fig ) , we have information from six different biological perturbation experiments ( S1 Table ) . It is known that ind expression is abolished in DL mutants [34] and becomes weaker in ZLD mutants [35] . Its peak expression reduces to ~50% of its wild-type level upon mutation of the four strongest ZLD binding sites [21] . ( We call this experiment ‘ZLD site mut . ’ . ) Removal of the strongest DL site ( ‘DL 1 site mut . ’ ) has no observable effects on the expression [36] and removing three overlapping DL sites ( ‘DL 3 site mut . ’ ) greatly diminishes peak expression [21] . Knockout of SNA ( experiment ‘SNA KO’ ) leaves ind expression unaltered [17] , while knocking out VND ( ‘VND KO’ ) causes the domain of expression to expand ventrally [37] , and CIC site mutagenesis ( ‘CIC site mut . ’ ) expands ind expression dorsally [38] . We evaluated each of the six perturbation experiments ( two TF knockouts and four site mutagenesis experiments ) using the approach introduced in the previous section–begin with the ensemble of models that explain wild-type gene expression , construct a filtered ensemble that additionally explains the perturbation results ( see Methods and S2 Fig ) , and calculate the difference in entropy ( ‘information gain’ ) . The values assigned to these experiments are shown in Fig 3A , and we note that the SNA and VND knockout experiments were the most informative in this group . Evaluating a new experiment , in our scheme , involves ruling out from the original ensemble a subset of models inconsistent with that new experiment . Recall that models in the ensemble were clustered , with the informal understanding that each cluster represents a distinct mechanistic hypothesis . Thus , if an entire cluster is ruled out by a particular experiment , one may interpret it as ruling out a particular mechanistic hypothesis . Table 1 shows the sizes of clusters in the original ( wild-type ) ensemble of models and the effect of filtering with each perturbation experiment . We note that an experiment ( ‘DL 3 site mut . ’ in Table 1 ) may remove just one cluster , while retaining other clusters of models as feasible . There may also be experiments ( ‘SNA KO’ and ‘VND KO’ in Table 1 ) that rule out the majority of mechanistic hypotheses , retaining only 2–3 of the original clusters . The other scenario–where all clusters are retained but rendered substantially sparser–is also seen , indicating that the information gained by those experiments was more along the lines of quantitative refinement rather than qualitative pruning of the space of possible mechanisms . Fig 3B shows the above information theoretic evaluation of each experiment , compared to a simpler scoring scheme where entropy of an ensemble is simply the logarithm of the size of that ensemble , i . e . , where we assume a uniform discrete distribution on models . As expected , the two scores are highly correlated . Note that experiments were assigned values above under the assumption that they were the sole ( or first ) perturbation experiment performed . In reality , of course , a line of enquiry proceeds via a series of such experiments , begging the question whether a perturbation experiment can be informative on its own but not so much if it follows another perturbation experiment . We explored this question further , by examining every possible pair of experiments ( performed sequentially ) , and noted that there are indeed such examples . However , in the interest of continuity we do not discuss this analysis here , referring the interested reader to S3 Table . We next moved beyond asking ‘how much’ information was gained from an experiment to the more subjective question of ‘what’ information was gained . To answer this , it seems natural to compare the original ( wild-type ) ensemble of models to the filtered ensemble that is additionally consistent with the new experiment’s results . The challenge then becomes: how do we compare these two ensembles in a language that appeals to the biologist’s intuition ? One pragmatic approach that we devised , and illustrate here , is to identify a second experiment for which the two ensembles make markedly different predictions , and use this difference to illustrate the distinction between ensembles . For instance , consider the ‘CIC site mut . ’ experiment , which we saw above to be of modest information theoretic value ( Fig 3A ) . We also noted in Table 1 that this experiment induces a filtered ensemble with two of the eight original clusters completely ruled out and two additional clusters drastically reduced in size ( from ~900 and ~700 models to 2 and 1 models respectively ) , suggesting that certain plausible mechanistic hypotheses were indeed ruled out by it . To interpret this further , we considered the predictions of this filtered ensemble on the ‘DL 1 site mut . ’ experiment ( Fig 3E ) and found these to be in fair agreement with the true results from the literature [36] ( Fig 3C ) . We then noted that the wild-type ensemble , not filtered by the ‘CIC site mut . ’ experiment , is far more uncertain in its predictions about the ‘DL 1 site mut . ’ experiment ( Fig 3D ) . Thus , the ‘CIC site mut . ’ experiment informs us , correctly , that mutagenizing the strongest DL site in the enhancer should not result in a significant reduction in peak ind levels , a point that was ambiguous in the original ensemble . A similar approach can be adopted to interpret the information provided by other perturbation experiments . In our second example , we interpreted the ‘DL 1 site mut . ’ experiment by examining the predictions of its filtered ensemble on the ‘CIC site mut . ’ experiment , which according to the literature [38] shows an extension of the dorsal boundary of ind expression ( Fig 3F ) This derepression effect is much more accurately predicted by the filtered ensemble ( Fig 3H ) , while the original ensemble’s average prediction is less definitive in predicting this effect ( Fig 3G ) . In other words , the ‘DL 1 site mut . ’ experiment informs us that CIC is an important repressor of the ind gene , setting up its precise dorsal boundary . For our third example , we note that the filtered ensemble of the ‘VND KO’ experiment accurately predicts that a genetic knockout of SNA will not affect the ventral boundary of ind expression ( Fig 3I and 3K ) , while the original ensemble erroneously predicts ventral de-repression ( Fig 3J ) . In other words , the ‘VND KO’ correctly informs us that SNA does not position the ventral boundary of ind expression . Thus , these three examples show how the information gained by an experiment can be interpreted by examining unique aspects of predictions of that experiment’s filtered ensemble on a second experiment . Similar to ind , single minded ( sim ) is dorso-ventral patterning gene in D . melanogaster that has been the subject of many biological experiments that describe the regulators of the gene , delineate its enhancer [39–41] , and characterize the combinatorial action of multiple TFs and cell signaling in the formation of the precise expression pattern driven by the sim enhancer [42] . The sim gene is initially expressed at the cellular blastoderm stage in a narrow row of width equal to two cells along the dorso-ventral axis at the mesectoderm ( the boundary between mesoderm and neural ectoderm ) [40 , 41] ( Fig 4A ) . Sim acts as a master regulator during the development of central nervous system ( CNS ) [43] and the confinement of its expression to the narrow line of cells is essential for the formation of the ventral midline and CNS during gastrulation [39 , 44] . This precise pattern of expression can be explained by a complex regulatory mechanism that involves Notch signaling [45–47] . On the ventral side , DL and Twist ( TWI ) activate but SNA represses the expression in the mesoderm [39 , 44] . Expression on the dorsal side is inhibited directly by Suppressor of hairless ( Su ( H ) ) , which is the only known repressor of sim in the neuroectoderm [45] , but is believed to have an activating influence on sim in the mesoderm region [45 , 48 , 49] . The sharp dorsal boundary of sim is formed because Notch signaling converts the ubiquitously expressed Su ( H ) from a repressor in dorsal regions to an activator in ventral regions exactly at the mesectoderm [37 , 45 , 48 , 50] . With these pieces of mechanistic information in hand , we employed GEMSTAT to model the expression driven by the sim enhancer . Then , we used the procedure introduced above to examine different perturbation experiments related to this enhancer reported in the literature , and quantify and interpret the ‘value’ of these experiments , after the fact . To our knowledge , this work is the first attempt to computationally model the expression of the sim enhancer , using the combinatorial action of TFs and signaling [45 , 47 , 51] . We built an ensemble of models that predicts the wild-type expression profile of sim accurately ( Fig 4B and Methods ) from its wild-type enhancer ( ‘2 . 8sim’ ) . We then considered several experiments reported in the literature pertaining to this gene , with the goal of computing the information gain from each experiment and interpreting the information they provide . Each of the nine experiments considered is a reporter assay with a variant of the wild-type sim enhancer , and we used its observed readout to construct objective criteria ( S2 Table ) for filtering models and creating a ‘filtered ensemble’ for that experiment ( S6 Fig ) . This allowed us to quantify the information gain score of each experiment , using the procedure described in previous sections ( Fig 4C and 4D ) . This revealed that the experiment ‘2 . 8simΔSD16’ , representing a deletion of two segments ( harboring a SNA site and an E-box element respectively ) from the wild-type enhancer 2 . 8sim , is the most informative ( value 2 . 25 ) , while seven of the other eight experiments are substantially less informative ( about 0 . 5 or less ) . Following the procedure of the previous section , we then sought to interpret the information gained by this experiment . This was most apparent when we used the filtered ensemble of this experiment to predict the outcome of another experiment ( ‘mesectoderm2 . 2’ ) . This second experiment is the reporter readout of a 2 . 2-kb sequence upstream of the early sim promoter and overlapping with the wild type enhancer 2 . 8sim considered above . According to the literature [52] , the expression driven by this sequence is unchanged ( Fig 4E ) from wild-type . The filtered ensemble of the ‘2 . 8simΔSD16’ experiment [39] can predict this known outcome accurately ( Fig 4G ) , while the wild-type ensemble is far more uncertain in its prediction ( Fig 4F ) . The ‘2 . 8simΔSD16’ experiment tests the effect of deletion of SNA sites on the gene expression and restricting the wild-type ensemble based on results of this experiment informs us which group of SNA sites is important to set up the precise expression of the sim gene .
We used the GEMSTAT model from [22] with 13 different parameters , including two parameters for each of the five TFs: one pertaining to TF-DNA binding ( KTF ) and one to the TF’s effect on transcription rate ( αTF ) . Moreover , the model has one parameter for DL-ZLD cooperativity ( wDL−Zld ) , and another parameter reflecting the baseline transcriptional rate ( qBTM ) . The repressor CIC has a uniform dorso-ventral expression profile but its repressive effect is attenuated in the neuroectodermal region by locally activated ERK , through a reduction in CIC-DNA binding; this effect , as modeled in [21] , is represented by a free parameter ( Cicatt ) . The expression pattern of each TF and the ind gene was scaled in the range of zero to one . Each expression profile is represented by a 50-dimensional vector ( S1A Fig ) , with dimensions corresponding to equally spaced bins along the D/V axis ( bin 1 = ventral end ) . The ind gene is expressed in only 5 to 7 of these bins ( bins 22–28 with the peak of expression at bin 25 ) . To construct the ensemble , we sampled models from the 13-dimensional parameter space following the procedure of Samee et al . [21] . We divided the range of each parameter into two halves ( using log scale for K parameters of all TFs , α parameters of repressors and for qBTM , and linear scale fo: α of all activators , wDL−Zld and Cicatt ) , and sampled 1000 points in each cell of the 13-dimensional space , for a total of ~8 million models . Such a dense sampling was possible because the number of parameters ( 13 ) is modest . Among these randomly generated models , we retained those with SSE ( sum of squared errors ) score between the real and predicted ind profile less than 10% . This higher initial threshold allows us to get good initial points . We then used GEMSTAT’s optimization routine to locally optimize the initial ensemble , following which we filtered for models that have SSE score less than 5% . This stricter error threshold was determined by a visual inspection of many examples , as one that allows the main spatial pattern of expression to be preserved in the predictions meeting that threshold ( see S1 Appendix for further details ) . We call the resulting collection the wild-type ensemble of models . It contains more than 5000 distinct parameter settings from about 600 different regions , and all of these models make good predictions on the wild-type data , as per visual inspection ( S1B Fig ) . Each model is represented by its parameter vector , and we first scaled the parameters using min-max scaling to place all the dimensions on the same scale of 0–1 . The models were then clustered based on a multivariate density estimation method , Mclust [53 , 54] , using its R implementation . This method approximates the complete collection of models as a ( generally non-uniform ) mixture of Gaussian distributions , each component Gaussian representing a cluster , with its own mean ( cluster center ) and covariance matrix ( S5 Fig ) , while simultaneously determining the optimal number of clusters . Next , we modeled each cluster C as a uniform mixture of nC = |C| Gaussian distributions with each of the nC models of the cluster as mean , and a common covariance matrix ΣC estimated for the cluster by the Mclust method . The following formula describes the probability distribution over the space of models θ: P ( θ ) =1N∑C=1N1nC∑i=1nCN ( θ;μiC , ΣC ) ( 1 ) where N is the number of clusters , C indexes these clusters , nC is the number of models in cluster C , μiC is the ith model in the cluster C , and ΣC is the covariance matrix of cluster C . Since models in the ensemble are not simply random samples of the probability distribution , but a collection of local optima obtained from initial random samples as seeds , we did not use a standard density estimation technique to build a density function . ( See Discussion for our choice of the above methodology as opposed to more sophisticated sampling techniques . ) We expected the distribution to reflect the fact that each model is a local optimum in the landscape of models scored by SSE measure and they group into several clusters . These clusters reflect different hypotheses for the biological mechanism of the underlying system and we desired that the probability density function put equal weights on them . Within each cluster , it is possible to observe several equally good local optima ( peaks of the SSE landscape ) ; we selected each such optimum as a local peak for the probability density function as well , constructing a Gaussian Mixture with a fixed covariance matrix to represent a cluster , with component Gaussians centered on the optimized models in that cluster . It is worth noting that our modeling of the desired probability distribution as a mixture of Gaussian distributions has implicit ties to the ‘MaxEnt principle’ , since the Gaussian is the maximum entropy distribution under a given mean vector and variance/covariance matrix . We calculated the entropy function using a discrete version of the probability density function in ( 1 ) . The discrete probability piC of the model i in the cluster C is set to be proportional to the value of the continuous probability density function at the location of the model . Each piC receives contributions from all Gaussians in the mixture model . We set the constant of proportionality such that for each cluster C we have ∑i=1nCpiC=1N where N is the number of clusters . We then estimated the entropy of this discrete probability density function using the formula H ( P ) =−∑C=1N∑incpiC×logpiC [55] . Suppose we are given an ensemble of models obtained from a set of experiments and have calculated its probability distribution . We are then given the results of a new experiment . This is typically in the form of gene expression level ( s ) driven by an enhancer sequence in one or more cellular contexts described by their TF concentration profiles . We assess if the predictions made by any model in the ensemble are consistent with these results , by comparing the model’s predictions to the observed gene expression levels through a single goodness-of-fit score such as SSE ( sum of squared error ) , and discard that model if this score is worse than a pre-determined threshold . Repeating this process for each model in the ensemble , we obtain a ‘filtered ensemble’ that is a subset of the original ensemble . We then modify the probability distribution over the filtered ensemble by ( a ) retaining the cluster assignments and cluster covariance matrices of the original ensemble , but ( b ) removing any cluster that has no remaining models and readjusting weights of remaining clusters to add to 1 , and ( c ) redefining the Gaussian Mixture Model for each cluster to have component Gaussian distributions centered at each model remaining in that cluster ( also with uniform weights ) . Having thus defined a probability distribution over the filtered ensemble , we calculate its entropy as above and assign the difference of entropy between the ( distributions over the ) original and filtered ensemble as the information theoretic ‘value’ of the new experiment . We constructed a one-dimensional vector to describe the expression readout of the sim enhancer along the D/V axis , as recorded in the literature . This vector ( and other expression vectors described here ) has 25 dimensions , representing equally spaced positions ( ‘bins’ ) along the axis from the ventral end to the mid-point of the axis . ( These are the same as bins 1–25 of the 50 bins considered in previous sections . ) The expression in each bin has a value between 0 and 1 that corresponds to the relative amount of gene expression observed in that bin . The expression profile of the wild-type sim enhancer was represented as a Gaussian curve that has its peak at the location where SNA expression changes from high to low ( 12th bin from ventral end ) , i . e . , at the known location of the sim expression peak . The variance of the Gaussian is set to be small enough that the ‘width’ of the expression profile is similar to the narrow domain in which SNA goes from high to low ( Fig 4A ) . We obtained TF ( protein ) expression profiles of DL , TWI , and SNA from Zinzen et al . [17] and represented them in the same 25-dimensional vector format as above ( Fig 4A ) . We considered DL and TWI as activators and SNA as a repressor . The other important regulatory input considered was Su ( H ) , which is a maternal protein uniformly expressed across the D/V axis . It is believed to be a repressor , but Notch signaling activated by the effect of SNA on Notch-Delta endocytosis switches the role of Su ( H ) from a repressor to an activator [45–47] in domains of SNA expression ( mesoderm ) . Since GEMSTAT does not allow for such a ‘role-switch’ for any TF , we separated the uniform expression profile of Su ( H ) into two separate profiles ( vectors ) , one for each role: an ‘activator Su ( H ) ’ with an expression profile similar to SNA but extended to include the mesectodermal positions and a ‘repressor Su ( H ) ’ with its complementary profile . In this manner we capture the prior knowledge of the ‘role-switch’ of Su ( H ) at the peak expression of sim . The sim enhancer sequence and TF motifs , required by GEMSTAT , were taken from Fly Factor Survey [56] .
Determining regulatory mechanisms shaping the spatio-temporal pattern of a gene of interest is a tedious process . While high throughput technologies provide helpful clues and narrow the space of possibilities , the ‘gold standards’ for demonstrating the regulatory influence of a transcription factor on a gene–a combination of TF knockout or overexpression ( and observed effects on gene expression ) , TF-DNA binding assays , site mutagenesis and rescue experiments–involve substantial investments . Guidance about the most insightful experiments to perform , given current knowledge about the gene’s regulation , can thus be highly beneficial . Typically , such choices are made by the biologist by relying on their intuition . We asked ourselves if the process of designing experiments to gain deeper understanding of a gene’s regulatory mechanisms may be made systematic . This immediately presented two major conceptual challenge: first , how do we formalize what is ‘current knowledge’ about the gene’s regulation , and second , how do we measure how insightful or informative an experiment is ? Answers to these questions appear to be necessary before we could systematize the process of experiment selection or design , mentioned above . In this manuscript , we take present a possible solution to these challenging problems by making use of a previously established quantitative modeling framework that relates trans- and cis-regulatory information to gene expression levels , and combining the framework with ensemble modeling and information theoretic ideas . In the future , our approach can be combined with well-established ideas in statistical experiment design [28 , 29] to develop a full-fledged formal approach to investigation of gene regulatory mechanisms . We approached the goal of formalizing current knowledge about a gene’s regulation by using the GEMSTAT framework of gene expression modeling . ( Other related models , e . g . , [13] , would also have been similarly usable . ) Here , current knowledge of a gene’s enhancer sequence ( s ) and its known regulators ( TFs ) is encoded into a mathematical function that is consistent with data on the gene’s and TFs’ expression levels in multiple conditions or cell types . This not only forces the qualitative knowledge of regulators into a precise quantitative form , it also explicitly captures complexities and subtleties associated with combinatorial action of multiple TFs . Furthermore , the model has free parameters representing important but often uncharacterized biochemical properties of the regulators , viz . , free energy of DNA binding and strength of regulatory influence , and the modeling step involves assigning values to these parameters so as to match available data . In this step , one is often faced with many distinct parameter settings that appear equally plausible in light of available data , and these different parameterizations represent ambiguities in current mechanistic understanding of the gene’s regulation , even when the likely regulators are qualitatively characterized . In our approach , such ambiguities are explicitly catalogued in the form of an ensemble of models consistent with data . To address the other conceptual challenge mentioned above ( ‘how informative is an experiment ? ’ ) , we compared the ensemble representing prior knowledge/data to that representing new experimental data in addition to the prior information . It was natural to consider using the information theoretic ideas of entropy and information gain for purposes of this comparison . We therefore devised an approach to define a probability distribution over models in the ensemble , and to estimate the entropy of the distribution; the information gain was then defined as the difference in entropy of the two ensembles . We demonstrated the use of our approach in the context of two genes in early fruitfly development–ind and sim–whose regulatory mechanisms have been studied through several perturbation experiments ( TF knockouts , site mutagenesis , variant enhancers , etc . ) reported in the literature . In each case , we started with the wild-type enhancer and likely regulators as ‘current knowledge’ and ( retroactively ) quantified how informative each of the perturbation experiments is . We also presented objective observations about each experiment that suggest the specific insights it added to our understanding of the gene’s regulatory mechanisms . In the case of ind , we additionally applied our experiment-scoring framework in a more controlled , semi-synthetic setting , where real data on the gene were used to first select a unique model as the underlying ‘truth’ , and used to provide the results of in silico perturbation experiments . We note however that the information gain values computed by our method are not comparable across different studies , e . g . , between the ind and sim studies considered here; they are only comparable across different experiments for the same gene , when evaluating those experiments for additional insights over a common set of current data/knowledge . Methodologically , an important feature of our approach was the generation of ensemble by uniform sampling in the multi-dimensional space , followed by optimization , as was done in [21] . Alternative sampling algorithms can be used to generate a large sample from optimal regions of the parameter space . Sampling methods such as Bayesian optimization techniques can be used to efficiently search for optimal parameters of any given model with nonlinear cost functions [57] . However , in practice these sampled optimal solutions may not be representative of every possible region of the parameter space with similar goodness of fit . One example of such a Bayesian optimization algorithm is the spearmint package . Spearmint , when applied to our problem , produced only a few optimal models . Building a large ensemble of models that represents every locally optimal region requires running the method multiple times , which is very slow , due to slow convergence time [58] . The spearmint method yields only few data points as the result of optimization and they are usually close to each other . Unless we perform a detailed sampling around those points or combine such techniques with a more global sampling approach , we do not have a diverse ensemble to work with . On the other hand , since the number of parameters in our model is small ( less than 20 ) , we could afford to do a dense uniform sampling of the parameter space with our approach . Our ensemble generation process has ties to Bayesian inference , as it constructs a distribution over models M , given data D . The more common approach to this is to sample from the posterior distribution P ( M|D ) , using a suitable likelihood model . There are two reasons why we chose not to do this in our approach . First , we wished to impose upon the distribution the property that different high-density regions of the space have equal relative weights ( probability mass ) , since each of these regions represents a distinct mechanistic hypothesis to us . This property is technically challenging to encode in the form of a prior distribution , and would require substantial research into the Bayesian inference methodology , which was not our main focus . Second , we found that a standard Bayesian optimization technique ( which we tested ) was not very efficient at sampling the parameter space globally . We believe that the framework established in this work can be used in future work to formalize experiment design strategies for gene regulation studies . For instance , we may compute the expected information gain [59] of various possible future experiments and select the best one . Given a candidate future experiment , we may first use each in the current ensemble to predict its outcome , compute the resulting information gain , and then compute an expectation of this value over the entire ensemble , using the probability distribution introduced in our work . | In-depth studies of gene regulatory mechanisms employ a variety of experimental approaches such as identifying a gene’s enhancer ( s ) and testing its variants through reporter assays , followed by transcription factor mis-expression or knockouts , site mutagenesis , etc . The biologist is often faced with the challenging problem of selecting the ideal next experiment to perform so that its results provide novel mechanistic insights , and has to rely on their intuition about what is currently known on the topic and which experiments may add to that knowledge . We seek to make this intuition-based process more systematic , by borrowing ideas from the mature statistical field of experiment design . Towards this goal , we use the language of mathematical models to formally describe what is known about a gene’s regulatory mechanisms , and how an experiment’s results enhance that knowledge . We use information theoretic ideas to assign a ‘value’ to an experiment as well as explain objectively what is learned from that experiment . We demonstrate use of this novel approach on two extensively studied developmental genes in fruitfly . We expect our work to lead to systematic strategies for selecting the most informative experiments in a study of gene regulation . | [
"Abstract",
"Introduction",
"Results",
"Material",
"and",
"methods",
"Discussion"
]
| [
"information",
"entropy",
"ecology",
"and",
"environmental",
"sciences",
"gene",
"regulation",
"experimental",
"design",
"research",
"design",
"regulator",
"genes",
"probability",
"distribution",
"mathematics",
"gene",
"types",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"gene",
"expression",
"soil",
"science",
"soil",
"perturbation",
"probability",
"theory",
"mutagenesis",
"genetics",
"information",
"theory",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences"
]
| 2018 | An information theoretic treatment of sequence-to-expression modeling |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.